High Performance Work Practices and Firm Training

※ This is a revised version of the paper presented at the KLI/ ILO Tripartite Workshop on Skill

Development, High Performance Work Organization and Social Dialogue, held in Seoul, Korea, on

March 6, 2003.

High Performance Work Practices and Firm Training

Dongbae KimResearch Fellow

Korea Labor Institutee-mail: [email protected]

April 2003

1. Introduction

Advances in IT technology gave rise to new industries while bringing changes

to existing businesses, intensifying global competition, shifting the source of

comparative advantage, and ultimately resulting in globalization. Growing importance

is being placed on human resources as a determinant of corporate competitiveness.

Competition among corporations in the 21st century is shifting from that based on

capital or tangible resources to one dominated by human resources (Pfeffer, 1994).

In order for a company to gain a competitive edge, it must continuously invest

in workforce training to increase their knowledge and skills. At the same time, it needs

to inspire the employees to become committed to the organization. Several requisites

that are called for in order to achieve the organizational capacity to attain competitive

advantage include: accumulation of human capital through sustained training, sharing of

individual knowledge within the organization, accumulation and constant revision of

organizational routine to become a learning organization.

The work system that maximizes worker’s competence and organizational

commitment is referred to as a “high-performance/involvement work system.” The

high-performance work system is a non-Tayloristic or high-involvement work

organization, characterized by high degree of autonomy of work groups, high level of

task integration and employee participation in innovative activities at the workplace.

The high-performance work system is also supported by human resources

management(HRM) that reinforces workers’ role structure.

As indicated in its description as a system based on the autonomy of skilled

workers, the high-performance work system is closely related to firm training.

(MacDuffie & Kochan, 1995). Under this system, quality and maintenance tasks

previously performed by special departments are delegated to workers, and work groups

independently take charge of the task-related decision-making formerly undertaken by

supervisors or managers. The system is also marked by workplace participation in

which workers suggest ideas for problem-solving and continuous improvement.

Since high-performance work systems require workers to take on broader roles,

work competence as well as self-governing and problem-solving capabilities must

accompany the workers in order to effectively carry out their roles. Thus, high-

performance work systems also call for continuing upgrading of workforce skills. Based

on the premise that high-performance work systems require enhanced workforce skills,

this study will analyze the effect of high-performance work system on firm training and

make some policy suggestions.

2. Review of Literature and Hypotheses

2.1 High-performance work system

High-performance work systems have been an important topic for studies on

HRM and industrial relations ever since it was first mentioned in the report “America’s

Choice” of the National Commission on Skills of the Workplace in 1989 (Cappelli &

Newmark, 1999). This work system emerged in the 1980s when the U.S. was trying to

emulate the success of Japan. Key features of this system take after the work

organization and HRM practices of large Japanese corporations (Doeringer et al., 1998;

Cappelli & Newmark, 1999).

Critics of the term “high-performance” instead use such adjectives as

“innovative,” “transformational,” “alternative,” “flexible,” and “involvement.” The use

of the term “high-performance” may create confusion as to whether the term is merely

an expression of desire or actually a proven means to enhance performance. Moreover,

the term invites misunderstanding as any system, regardless of its characteristics, can be

deemed high-performance as long as it generates good performance records.

According to some researchers, the key to high-performance work systems lies

in worker participation in the decision making at the workplace (Cotton, 1993; Parks,

1995; Delaney, 1996). A high-performance work system entails a high-involvement

work organization that requires increased workers participation and roles in the

workplace, not a Taylorized organization that minimizes the role of workers. Instead of

employing HRM policies and practices designed to control workers, the high-

performance system employ high-commitment HRM practices that promote workers’

voluntary involvement and dedication as well as skill formation. Some researchers

separate non-Taylorized or high-involvement work organization as the core of high-

performance work systems and commitment-oriented HRM as a complement to work

organization (MacDuffie, 1995; Pil & MacDuffie, 1996).

High-involvement work organization can be further categorized into three

dimensions: task integration, autonomy of work groups, and workplace participation,

although the combination among these dimensions may differ by county or by company.

Task integration negates excessive specialization and reintegrates job-related tasks into

a single job. Indicators of task integration include workers’ responsibility for

maintenance or quality work, job rotation and task composition. Autonomy of work

groups relaxes vertical specialization of management hierarchy, allowing work groups

to independently decide on the planning and control functions usually reserved for

managers. Autonomy is assessed by whether autonomous work teams are in place or by

how much autonomy is retained by each work group. Workplace participation negates

the monopoly of problem-solving and improvement processes by engineers or managers

and its indicators are off-line activities like QC or worker suggestions.

Although there is no consensus on what constitutes commitment-oriented HRM

practices, most agree that workers should be recognized as important stakeholders just

as shareholders and workers’ interests should be faithfully represented. In addition,

HRM should be based on trust, humanistic philosophy and long-term perspective.

Recent studies also classify commitment-oriented or supporting HRM into the

dimensions of motivation, skill formation and empowerment or information sharing

(Appelbaum et al., 2000; Gardner et al., 2001; Wright & Boswell, 2002). The dimension

of motivation is comprised of such practices as life long employment, internal

promotion and profit sharing, and skill formation consists of worker training. Lastly,

empowerment or information sharing is built on communication, workplace

participation or information sharing practices. Although some researchers of HRM do

not set aside work organization as a separate subject but include it in empowerment, it

would be better to set up a separate dimension for work organization as some study call

for an area clearly distinct from HRM. High-performance work systems can be

illustrated as in Figure 1.

<Figure 1> High-performance Work System

Motivation

Skill formation

Informationsharing

High-involvement work organization

Supporting HRM

Task integration

Workplaceparticipation Autonomy

2.2 High-performance work systems and firm training

Studies on the relationship between high-performance work systems and firm

training follow two broad trends. One examines firm training as an element of the high-

performance work systems and the other studies the effect of high-performance work

systems, excluding the element of firm training, on firm training. Most studies on high-

performance work systems consider skills formation as a crucial element, and thus

examine what kind of effect the work system as a whole has on firm performance or

workers. As seen in Figure 1, high-performance work systems consist of non-Taylorized

or high-involvement work organization and supporting HRM, under which skills

formation takes up a dimension.

It was in the mid-1990s when high-performance work systems and firm training

began to be thought of as separate concepts and the effect of the former on the latter

came under scrutiny (MacDuffie & Kochan, 1995; Osterman, 1995; Wagar, 1997;

Lynch & Black, 1998; Frazis et al., 2000; Whitfield, 2000).

Figure 1 classified training as an element of high-performance work systems.

However, system elements may be loosely connected so that the internal consistency

among elements is not automatically guaranteed. For instance, the adoption of high

involvement work organization may not lead to the full-fledged introduction of

supporting HRM, including skills formation, and vice versa. Exploring the relationship

between high-performance work systems and skill formation as a separate dimension is

necessary in order to verify the internal consistency of the system.

Increased interests in examining separately the high-performance work systems

and skill formation stemmed from the need to increase investment in firm training in

order to enhance competitiveness. However, since corporate investment in training is

usually determined by the characteristics of work system, companies eventually face a

practical matter of introducing high-performance work systems to increase investment

in training (Osterman, 1995; MacDuffie & Kochan, 1995). In short, as high-

performance work systems create the demand for high-skilled workers, skill-related

government policies should also change in order to encourage the introduction and

dissemination of high-performance work systems that stimulate such demand.

<Table 1> High-performance work system and firm trainingData Measurement of

trainingMeasurement of

work system Result

MacDuffie &Kochan(1995)

AutomobileAssembly plant(N=57)

Hours of training(Off-JT & OJT)

Team, off-line team,suggestion, jobrotation,responsibility forquality, selectioncriteria, profitsharing etc.

Positive

Osterman(1995)

U.S.A.(N=878)

Proportion of coreworkers trained

Team, TQM, QC,SPC, job rotation,problem-solvinggroup

Positive

Wagar(1997) Canada(N=569)

Num. ofoccupational groupsreceived training ,proportion ofworked trained

Team, QC, QWL,problem-solvinggroup

Positive

Lynch &Black(1998)

U.S.A.(Educational Qualityof WorkplaceNational EmployerSurvey)

Proportion ofworkers trained,types of training

Team, TQM, jobrotation,Benchmarking,number ofHierarchy, span ofcontrol

Benchmarking,TQM, team(positive)

Fraziset al.(2000)

U.S.A.(N=. 1,062)

Training incidence,hours of training,trainingexpenditures

Team, QC, TQM,job rotation, skill-based pay,workplaceparticipation, jobredesign, re-engineering, JIT,peer-rating

Positive

Whitfield(2000)

U. K.(N=657)

Proportion ofworkerstrained*days spentin training

Team, QC, teambriefing, flexibleassignment

QC, team briefing,flexibleassignment(positive)

From a theoretical point of view, the studies on the effect of high-performance

work system on firm training have opened a new horizon in the debates on

skills(Osterman, 1995). While the debates on skills gave rise to a series of trends,

including case studies since Braverman, studies using the DOT, and researches using

corporate information, the debate on “de-skilling” versus “up-skilling” is not over. Thus,

it is worthwhile to identify which work system increase investment in training since it

looks at training in terms of corporate demands. Table 1 shows the summarized review

of literature on high-performance work systems and firm training.

As seen in Table 1, although previous studies on the relationship between high-

performance work systems and firm training focused on the U.S., the U.K., Canada and

other countries, and employed different measurement of high-performance work system

and training indicators, they all resulted in similar outcomes. In sum, high-performance

work system and its elements are factors that demand an increased firm training.

2.3 Hypotheses

As the saying “you leave your head in the dressing room and take only your

hands and feet to the workplace” implies, Taylorized work organizations minimize the

role of workers at the workplace. On the contrary, in high-involvement work

organizations, quality and maintenance tasks previously supervised by specialized

departments are now under the responsibility of workers. In addition, work groups are

in charge of work-related decision-making, which used to be under the exclusive control

of supervisors and managers. Workers also propose ideas and take part in a range of

activities concerning problem-solving and improvements(H. Kim & D. Kim, 2001).

Since high-involvement work organization requires much more input from

workers, workers need to develop their capabilities in order to carry out their roles

effectively. To take charge of maintenance and quality work, for instance, workers need

training in relevant fields as well as multifunctional training to enable flexible work

allocation like job rotation. Furthermore, members of work groups should be provided

training on decision-making process, communication and conflict management, and

managerial work.

If high-involvement work organization is a role structure of workers,

supporting HRM plays a complementary role that provides competence, motivation and

resources (such as information) needed to put such work organization in place. Because

of the complementarity between work organization and HRM, the more deeply

established the involvement-oriented work organization, the more likely for the

supporting HRM comprised of motivation, information sharing and skill formation to be

adopted.

As seen in the HR philosophy of human investment model, the principle of

supporting HRM is grounded on investment in workers. For example, motivation

consists of employment guarantee, relatively high wage and internal promotion, and

information sharing has to do with sharing business and task-related information with

workers. Skill formation, which we categorized as a separate dimension, signifies direct

investment in workers. There is also complementarity among the dimensions of

supporting HRM, with higher motivation or information sharing leading to more

investment in skill formation. Such reasoning supports the hypotheses concerning

corporate investment return in the theories of human capital or internal labor market,

which predicted HRM encouraging long-term employment would increase investment

in firm training (Frazis et al., 2000).

Studies on high-performance work systems focus on the effect of a “system”

of interconnected practices rather than individual practices, given the complementarity

of elements. Complementarity among work practices is observed, for instance, when

performance improvement gained from adopting two practices together is greater than

the sum of improvements gained from adopting the two practices individually (Pil &

MacDuffie, 1996). When complementarity among elements exists, the elements

combined as a system create a greater synergy effect than the sum of individual effects.

Therefore the following hypotheses can be derived from the above discussion.

H1: High-performance work system would increase firm training.

H 1-1: High-involvement work organization would increase firm training.

H 1-2: Supporting HRM would increase firm training.

The variables that could possibly impact corporate investment in training other

than high-performance work systems, such as the firm age, industry, listings on the

stock market, labor unions, competitive pressure, competitive strategy and introduction

of Six-Sigma, have been controlled.

The younger the company, the more likely it is for it to invest in training to

catch up with the skill level of existing companies. In a study by Whitfield (2000), for

example, the older the company, the less it invested in training. However, social

structure and practices of the time tend to be imprinted on the organization at the time of

its founding. The relationship between the age of an organization and training may

differ, depending on the social and institutional environment at the time of the

establishment, discouraging such a definite prediction of the impact of years in

operation.

It is generally predicted that the larger the company, the greater the investment

in training (Jang-soo Ryu, 1997), because larger companies tend to enjoy more slack to

invest in training and the economy of scale related to training. Moreover, growth of

organization is accompanied by increased specialization, which brings about a greater

need for monitoring in order to coordinate the work efficiently. The cost of monitoring

increases concurrently with size, which makes training more appealing than costly

monitoring. In short, the management would find it more tempting to augment

investment in training if internalized monitoring through training turns out to be more

effective than direct monitoring (Scott & Meyer, 1994). However, a word to the wise in

this case is that the relationship between firm size and training investment may be either

linear or non-linear (Knoke & Kalleberg, 1994). Previous studies by Wagar (1997),

Frazis et al. (2000), and Whitfield (2000) demonstrated a positive relationship, but those

by Knoke & Kalleberg (1994), Osterman (1995), Felstead & Green (1996) and J. Ryu

(1997) found no significant relationship.

High capital intensity is a characteristic of heavy industries. When capital

intensity is high, discretionary actions of workers substantially affect performance and

make it harder to monitor worker activities. Therefore, it may be more efficient to rely

on norms through training and autonomy of skilled workers. Heavy industries, therefore,

would invest in training more heavily than light industries. Multivariate analyses

indicated that result controlling 3-digit manufacturing industry showed similar outcome

compared with heavy industry dummy variable without controlling for 3-digit industry

dummy variables.

Listed companies or KOSDAQ-registered public corporations may also show

different levels of investment in training as compared to unlisted companies. Public

corporations are considerably exposed to governmental or public monitoring, thus very

likely to implement mandatory training. If the overall social atmosphere appears to

favor firm training, public companies would probably feel the normative pressure. On

the other hand, although the stock market may apply different degrees of pressure on

companies, public corporations may invest little in training, because the stock market is

generally skeptical about substantially uncertain investment in training.

Given our keen interest in the skill formation projects led by the labor-

management partnership, the relationship between labor unions and training investment

is a crucial research subject. However, both theoretical and experimental studies on the

effects of labor unions on corporate investment in training demonstrate a variety of

contradicting arguments and research findings. Labor unions may enhance investment

in training, for they reduce the turnover rate of workers by providing a collective voice

rather than exit. On the other hand, however, very likely promotion of unqualified

candidates allowed under the seniority-based scheme of labor unions will cut down on

firms’ incentive to invest in training (Freeman & Medoff, 1984; Knoke & Kallberg,

1994). In the meantime, Smith & Dowling (2001) hypothesized that the degree of

union-employer partnership and corporate investment in training are in an inverse U-

shaped relationship. This hypothesis is based on the reasoning that when the degree of

partnership between the labor and management is either very low or very high, attention

is placed more on the overall relationship rather than on the training of individual

workers.

Osterman (1995) found a significant positive correlation between labor unions

and firm training. However, Ryu (1997) in Korea and Whitfield (2000) in the U.K.

detected no significant relationship among the variables. It appears that labor union’s

policy on training is more important than the presence of unions itself, and the union’s

training policy is closely linked to the overall labor-management history of a given

country and the current environment.

The variable of product market is also a vital factor in training investment.

Fiercer competition in the product market could lead to increased investment in training

to strengthen organizational capability. As Keenoy (1995) and Osterman (1994) pointed

out, however, the reverse could be observed when control-oriented management style

takes hold in response to intense competition. Therefore, competitive strategies, rather

than the mere intensification of competition, has more impact on training investment.

Researchers maintain that the introduction of high-performance work systems in the

U.S. is greatly influenced by quality strategies (Cappelli et al., 1997; Lawler et al.,

1998). Furthermore, according to strategic HRM, investment in training is more likely

to jump when the management steadily pursues high road strategies based on quality,

diversity and speed, rather than low-cost strategy.

Six-Sigma is a recently adopted quality management tool. When Six-Sigma is

introduced as a quality management tool, training concerning quality management will

expand. A study by Felstead & Green (1996) could provide explanation for the effect of

Six-Sigma on training. They found that the acquisition of BS 5750 or ISO certificates,

outside control and changes in training features explained the unexpected phenomenon

of increased training in the companies suffering from a recession-induced fall in

revenue (N=27). Acquisition or maintenance of industry standards requires investment

in training and some companies or clients request standard certificates as a condition of

transaction. Therefore, companies that obtained such standard certificates can inject a

great sum of money in training. Six Sigma is expected to generate similar effect the one

observed in Felstead & Green’s (1996) study on standard acquisition.

3. Data and Measurement

3.1 Data

This study is based on the data of the Workplace Panel Survey (WPS)

conducted in 2002 by the Korea Labor Institute and utilizes training-related data from

the employment insurance database and corporate financial data from NICE Credit

Bureau.

The WPS data was analyzed by combining questionnaires for HR managers

(N=1,395) and labor affairs. The questionnaires for labor affairs were filled out by labor

affairs managers at workplaces together with worker representatives. This study first

matched HR managers response with the labor affairs questionnaires (N=1,245), and

only the representatives’ responses were included (N=73) from the workplaces where

labor affairs managers were not available for the survey. The remaining 77 workplaces

submitted responses only from HR managers.

Subjects of this study were limited to manufacturers, because, except for a few

customized survey questions, the tools to measure the work systems of non-

manufacturing industries were not fully developed, and the same limitation existed in

the data used in this study. Actually Cappelli & Newmark (2001) revealed that high-

performance work practices had completely different meanings for the manufacturing

and service industries. So, this study is confined to manufacturers because the study

employed the conventional work system measurement items developed by the

manufacturing industry. Originally there were 691 respondent manufacturers, but only

598 manufacturers were included in the final analysis, as 35 did not answer the question

that asked for the number of production workers and 38 companies answered “0” to the

same question. Out of the 598 workplaces, 529 included labor affairs management

questionnaires from labor affairs managers and 31 from worker representatives.

Therefore, the valid sample size was 560 workplaces, and the sample used in the final

analysis, excluding missing observations, totaled 554.

In order to add information on training investment, data on the 2001 Vocational

Capability Development Training were extracted from the employment insurance

database. The extracted data included the amount of insurance refund generated by the

implementation of the 2001 Vocational Capability Development Training(including paid

training leave) and the annual count of beneficiaries. Since the sample population of the

WPS coincided with that of the employment insurance database, matching of two sets

of data did not create any problems other than few minor adjustments. In addition,

corporate financial data from NICE Credit Bureau were utilized to obtain another data

on training expenditure.

3.2 Measurement

3.2.1 Dependent variables

Measurement of training investment varies by researchers. Osterman (1995)

measured the percentage of core workers receiving official training, and MacDuffie &

Kochan (1995) combined the OJT and Off-JT hours of newly recruited employees and

workers with more than one year of job experience. Ryu (1997) calculated the

proportion of official training investment by adding the training investment costs from

the balance sheets and manufacturing cost invoices. The problem with this measurement,

however, is that training cost is not a mandatory item on the financial statement.

Whitfield (2000) estimated the training figures by multiplying the percentage of

workers who completed official training courses over the last year by the hours. Frazis

et al. (2000) used three measurements – provision of official training, training hours and

training costs.

It is a well-known fact that accurate data on training investment is quite lacking.

Given the absence of reliable data, this study quantified the degree of training

investment into training expenditure, training coverage and annual training hours by

using different data sources. The training expenditure and the coverage data were

derived from the employment insurance database. Insurance premium refunded through

the 2001 Vocational Capability Development Training, including the paid training leave,

and the number of annual beneficiaries were divided by the total number of workplace

employees at the beginning of 2002.

Training investment data from the insurance database could also be limited, so

the training investment amounts recorded in the corporate financial statements obtained

from the NICE Credit Bureau were divided by the total number of employees at firm in

early 2002 to arrive at the per capita training expenditure. This variable is named

training expenditures1(See Table 2). However, this data is also problematic in that

financial data can be disclosed only to outside auditors and even they very seldom

record the amount of training investment in both the balance sheet and the

manufacturing cost invoice, substantially shrinking the number of usable cases.

Moreover, training investment is not a mandatory item on the financial statement and

could be classified under a different item to suit the circumstance of each company, so

the data does not accurately reflect the status of training investment. Therefore the

training expenditure1 obtained from the financial statements was used for the sole

purpose of comparing the result with that gained from the employment insurance

database.

Lastly, hours of training per person was created by using the number of the Off-

JT and OJT beneficiaries in 2001 and the yearly per capita training duration. This data

is drived from WPS. Hours of training index was calculated by dividing the number of

workers that received the Off-JT in 2001 by the number of employees in early 2002, and

then multiplying it by the day of training. The same calculation was applied to obtain

the duration of the OJT and finally hours of training calculated by summing up both

Off-JT and OJT.

The distribution of investment variables was skewed so the logarithmic series

was used in analysis. There were many instances where training investment amount and

beneficiary rate were “0” so the value of 1 was added to each figure and then the

logarithm was taken. In this case, if the value of training investment originally “0,” it

would still retain the value of “0” even after the transformation.

3.2.2 Independent variables

Supporting HRM was measured with motivation and information sharing. In

measuring motivation, 5-point scales were used to identify the strictness of selection

and the wage level, and the value of 1 was given when HR merit rating system and at

least one out of profit sharing or other group performance pay or employee stare

ownership were adopted. One point was also allotted when employment was guaranteed

by refraining from employee downsizing, such as layoffs or early retirement, after the

economic crisis. Motivation was quantified by combining these standardized variables.

Information sharing was measured by adding up the indices for the presence of business

briefings, newsletters with management information, hotline between the workers and

the management, and team briefing. The addition of standardized values of motivation

and information sharing items made up the supporting HRM index. The internal

consistency of the HRM index was α=.6124.

High-involvement work organization was measured by averaging the indices of

work group autonomy, task integration and workplace participation. Autonomy of work

group is a composite index of dummy variables, which were given one point when

answers were either “somewhat autonomous” or “completely autonomous” in regards to

work fare, work method, and work speed(each on a 5-point scale). Task integration was

obtained by adding one dummy variable, which was given the value of 1 when

production workers are responsible for quality, and another dummy variable that earned

one point when production workers were rotated. Workplace participation consisted of

variables that each received one point when the percentage of its production workers

participating in small group activities stood at 50% or more and when workers were

actively making suggestions. The standardized values of autonomy, task integration and

workplace participation were added up to obtain the work system index. The internal

consistency of the index registered α=.5963.

The age of workplace was calculated by subtracting the year of establishment

from 2002. The size of workplace was estimated from the number of employees and the

logarithm of it was used. Heavy industry dummy followed the categorization of Jeong

(1999), giving the value of 1 to manufacturing industry codes 23-24, 27, 28-35, and 371.

The dummy variable of labor union received one point when labor unions existed.

Competition intensification was identified by the factor score (α=.6122) of “the number

of competitors,” “changes in existing products or services,” “new product

development ,” “product demand,” and “the importance of quality” (5-point scale) in

the core product or services market during the last three years. The characteristics of

core products were categorized into low cost, quality, variety, speed and technological

edge (each on a 5-point scale) for the competitive strategy variable. These

characteristics were put through a factor analysis to yield two factors. Four items except

for low cost were classified as the first factor, which can be interpreted as high-road

strategy(α=.7352). Adoption of Six-Sigma was a dummy variable with the value of 1.

Descriptive statistical figures of these variables are presented in Table 2.

<Talbe 2> Means and Standard DeviationN Mean S.D.

Training Expenditures(1,000 Won) 552 36.06 237.54Percentage of Workers Trained 552 0.27 1.35Hours of Training(day) 498 4.43 10.45Training Expenditures1(1,000 Won) 189 237.70 343.06Work System Index 384 0.03 0.48

HRM Index(α =.6124) 468 0.01 0.49Motivation 477 0.01 0.49

Selective hiring 541 2.42 1.07Personnel rating 512 0.50 0.50Wage level 534 3.05 0.78Profit sharing 539 0.33 0.47Employment security 519 0.63 0.48

Information sharing 544 2.13 1.17Business briefing 545 0.66 0.47News letter 551 0.20 0.40Hot line 552 0.50 0.50Team briefing 552 0.78 0.42

Work Organization Index(α =.5963) 433 0.04 0.66Work Group Autonomy 517 1.05 1.19

Work fare 520 0.33 0.47Work method 519 0.36 0.48Work speed 520 0.37 0.48

Workplace Participation 516 0.77 0.77QC 521 0.22 0.42Suggestion 527 0.56 0.50

Task Integration 460 0.78 0.70Responsibility for quality 466 0.52 0.50Job rotation 532 0.23 0.42

Establishment Age 543 20.92 14.70Num. of Employee 554 565.18 2797.80Heavy Industry 554 0.59 0.49Listed Company 554 0.21 0.41Trade Union 554 0.37 0.48Competitive Pressure(α =.6122) 515 0.00 1.00High-road Strategy(α =.7352) 518 0.00 1.00Six-Sigma 533 0.17 0.38

4. Results

4.1 The status of investment in training

The descriptive statistics in Table 2 and correlations in Table 3 reveal the status

of how much investment is being made in training.

Under the assumption that no refund of insurance premium implies non-

implementation of paid training leave and other vocational capability development

training, 369 (66.85%) out of 552 companies were found to have invested in training in

2001 and the rest did not conduct any training. Annual per capita training investment

estimated from the refund amounted to an average of 36,000 won.

The investment amount may be considerably underestimated than the actual

training investment made by the companies, because, even when the companies poured

money into training, they were omitted from the analysis if employees did not get their

employment insurance refund for various reasons or if grants or other types of

investment were not included in the training investment category. For instance, Table 2

shows the annual per capita training expenditures1 in the corporate financial statements

standing at 237,000 won. Therefore, if we base our analysis on the financial statement

figures, the amount of employment insurance refund accounts for only 15.17% of the

actual corporate investment in training.

The annual benefit receipt rate of paid training leave and other vocational

capability development training posted 27%. The training coveragee in the employment

insurance database is derived from the annual number of beneficiaries, which could lead

to one person being counted twice, so the data should be interpreted with caution. The

annual hours of training per person measured by questionnaires was 4.43 days,

including Off-JT and OJT, which somewhat exceeded the findings of other studies in

Korea (Kim, 2000; Kim & Roh, 2002).

Three sources used to measure the level of investment in training each have

limits. Correlation among the training investment indices from three data sources (see

Appendix 1) all revealed positive relationships that were significant at 99% level,

except for the correlation between the hours of training obtained from questionnaires

and the training expenditures1 from financial statements. Therefore, each measurement

was found to have the minimum level of reliability, but it is difficult to identify which

data source most accurately reflects the degree of investment in training.

<Table 3> Correlation (N=336)1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

2 0.78

3 0.39 0.35

4 0.37 0.31 0.34

5 0.36 0.29 0.38 0.74

6 0.28 0.22 0.33 0.68 0.94

7 0.33 0.32 0.29 0.47 0.58 0.28

8 0.27 0.22 0.21 0.87 0.32 0.28 0.24

9 0.07 0.03 0.08 0.50 0.12 0.10 0.08 0.62

10 0.35 0.30 0.30 0.68 0.36 0.30 0.29 0.70 0.16

11 0.11 0.10 0.03 0.54 0.16 0.15 0.09 0.64 0.05 0.20

12 0.24 0.22 0.16 0.04 0.00 -0.04 0.08 0.06 0.04 0.11 -0.03

13 0.46 0.34 0.23 0.30 0.33 0.24 0.35 0.19 0.00 0.26 0.10 0.40

14 0.20 0.07 0.19 0.13 0.05 0.05 0.03 0.15 0.12 0.07 0.09 -0.11 0.04

15 0.22 0.17 0.15 -0.01 0.02 -0.01 0.07 -0.02 -0.02 0.00 -0.02 0.32 0.29 0.01

16 0.32 0.32 0.13 0.15 0.15 0.12 0.16 0.10 0.04 0.18 -0.03 0.42 0.51 -0.09 0.21

17 0.07 0.08 0.18 0.32 0.26 0.23 0.18 0.27 0.13 0.19 0.20 -0.01 0.12 0.06 0.08 -0.02

18 0.05 0.02 0.06 0.25 0.25 0.24 0.14 0.17 0.06 0.14 0.13 -0.04 0.12 -0.09 0.04 -0.04 0.37

19 0.25 0.19 0.15 0.25 0.20 0.14 0.24 0.20 0.07 0.24 0.09 -0.04 0.24 0.08 0.02 0.16 0.12 0.18

Notes: 1). Correlation greater than |.14| are significant at .01; r’s greater than |.11| are significant at .05. 2) 1=Training expenditures, 2=Percent of workers trained, 3=Days of training, 4=High-

performance work system, 5=Supporing HRM, 6=Motivation, 7=Infoamtion sharing, 8=High-involvement work organization, 9=Work group autonomy, 10=Workplace participation,11=Task integration, 12=Establishment age, 13=Establishment size, 14=Heavy industry,15=Listed company, 16=Trade Union,17=Competitive pressure, 18=High-road strategy, 19=Six-sigma

As shown in Table 3, the three indices of investment training are significantly

and positively correlated with: the high-performance work system indicator and its sub-

dimensional indicators of supporting HRM and high-involvement work organization;

motivation and information sharing, the sub-dimensions of high-involvement HR

management; and workplace participation from the sub-dimensions of high-

involvement work organization. The correlation coefficient of supporting HRM and

high-involvement work organization stood at 0.32 (p<.01), thus the effect of high-

involvement work organization without supporting HRM was also examined in the

multivariate analysis.

Contrary to the prediction that new companies would make a considerable

investment in training to catch up, the age of organization showed a positive

relationship with the amount of training investment. As projected, the correlation

between organization size and investment in training was positive one as was the

correlation between the dummy variables of heavy industry and Six-Sigma and training

investment.

Meanwhile, the unpredictable direction of the relationship between training

investment and two variables of public corporations and labor unions turned out to be a

positive one. The variables of competition intensification and competitive strategy show

positive yet very low correlation coefficient, while the correlation between these

variables and those of high-performance work system were quite high. Therefore, the

variables of competition intensification and competitive strategy appear to have fostered

corporate investment in training through high-performance work systems, which merit

future studies.

4.2 Factors influencing the degree of investment in training

Tobit analysis was used to estimate a coefficient since the training investment

indicator oftentimes recorded “0” when the amount of investment in training on the

financial statements was taken out. Hence, linear regression analysis (OLS) was used to

analyze training expenditures1, which is not affected by the value of “0.” (see appendix

2). Although not reported in this paper, OLS yielded results similar to those by Tobit

analysis.

Model 1 is the result high-performance work system index together with control

variables and Model 2 shows the result of injecting supporting HRM index and high-

involvement work organization index, the two sub-dimensions of high-performance

work systems. Model 3 presents the outcome of putting in only the work organization

index, taking into consideration the correlation between the supporting HRM and high-

involvement work organization. Model 4 had the sub-dimensions of supporting HRM

and high-involvement work organization instead of those indices themselves, and

Model 5 had only the sub-dimensions of high-involvement work organization, not those

of supporting HRM.

<Table 4> High-performance work system and Training Expenditure (Tobit)

M1 M2 M3 M4 M5

-2.510*** -2.383*** -2.496*** -3.058*** -2.781***Constants(0.534) (0.535) (0.508) (0.534) (0.506)0.012 0.013 0.006 0.012 0.005Age

(0.009) (0.009) (0.008) (0.009) (0.008)0.568*** 0.532*** 0.561*** 0.492*** 0.531***Size(0.110) (0.111) (0.106) (0.112) (0.107)

0.713*** 0.743*** 0.846*** 0.757*** 0.861***Heavy industry(0.231) (0.231) (0.223) (0.229) (0.222)0.501* 0.488* 0.546** 0.511* 0.589**Listed company(0.272) (0.271) (0.271) (0.268) (0.269)0.305 0.318 0.471* 0.310 0.466*Trade Union

(0.267) (0.266) (0.263) (0.264) (0.261)-0.126 -0.120 -0.062 -0.119 -0.059Competitive

Pressure (0.119) (0.118) (0.116) (0.117) (0.116)-0.093 -0.110 -0.036 -0.094 -0.028High-road strategy(0.125) (0.125) (0.122) (0.124) (0.121)

0.850*** 0.852*** 0.961*** 0.740** 0.878***Six-Sigma(0.293) (0.291) (0.288) (0.293) (0.288)

1.184***HPWS(0.263)

0.981***HRM(0.259)

0.608**Motivation(0.243)0.251**Information

sharing (0.106)0.374** 0.455***WO(0.181) (0.170)

0.024 0.053Work groupautonomy (0.093) (0.091)

0.440*** 0.496***Workplaceparticipation (0.154) (0.147)

0.013 0.021Taskintegration (0.156) (0.153)

Log L -589.43*** -587.91*** -672.47*** -584.78*** -669.65***N 358 358 401 358 401

Note: 1) Standard Errors in Parentheses 2) *p<.1, **p<.05, ***p<.01 (two-tailed)

<Table 5> High-performance work system and Training Coverage (Tobit)

M1 M2 M3 M4 M5

-0.523*** -0.508*** -0.514*** -0.643*** -0.569***Constants(0.101) (0.102) (0.094) (0.102) (0.094)0.001 0.001 0.000 0.001 0.000Age

(0.002) (0.002) (0.002) (0.002) (0.002)0.092*** 0.087*** 0.089*** 0.080*** 0.084***Size(0.021) (0.021) (0.020) (0.021) (0.020)0.039 0.043 0.062 0.045 0.064Heavy industry

(0.044) (0.044) (0.041) (0.044) (0.041)0.067 0.066 0.065 0.069 0.072Listed company

(0.051) (0.051) (0.050) (0.051) (0.050)0.093* 0.095* 0.110** 0.095* 0.110**Trade Union(0.050) (0.050) (0.048) (0.050) (0.048)-0.011 -0.010 -0.004 -0.011 -0.003Competitive

Pressure (0.023) (0.023) (0.022) (0.022) (0.022)-0.020 -0.022 -0.009 -0.019 -0.008High-road strategy(0.024) (0.024) (0.023) (0.024) (0.023)0.130** 0.130** 0.149*** 0.110** 0.137***Six-Sigma(0.055) (0.055) (0.053) (0.055) (0.053)

0.204***HPWS(0.050)

0.151***HRM(0.049)

0.084*Motivation(0.046)0.049**Information

sharing (0.020)0.075** 0.079**WO(0.034) (0.031)


0.069** 0.079***Workplaceparticipation (0.029) (0.027)

0.013 0.011Taskintegration (0.030) (0.028)

Log L -164.02*** -163.37*** -187.12*** -161.06*** -185.31***N 358 358 401 358 401


<Table 6> High-performance work system and hours of training (Tobit)

M1 M2 M3 M4 M5

-0.921** -0.718* -1.500*** -1.302*** -1.673***Constants(0.444) (0.434) (0.434) (0.440) (0.430)0.015** 0.018** 0.014** 0.017** 0.012*Age(0.007) (0.007) (0.007) (0.007) (0.007)0.111 0.049 0.240*** 0.022 0.208**Size

(0.094) (0.093) (0.092) (0.094) (0.091)0.643*** 0.692*** 0.627*** 0.715*** 0.652***Heavy industry(0.184) (0.180) (0.179) (0.178) (0.176)0.431** 0.422** 0.319 0.432** 0.360*Listed company(0.217) (0.211) (0.217) (0.207) (0.212)0.029 0.004 0.090 -0.033 0.052Trade Union

(0.216) (0.210) (0.213) (0.207) (0.209)0.165* 0.160* 0.238*** 0.168* 0.238***Competitive

Pressure (0.094) (0.092) (0.092) (0.090) (0.090)-0.072 -0.101 0.012 -0.083 0.024High-road strategy(0.101) (0.099) (0.099) (0.098) (0.096)0.244 0.270 0.219 0.170 0.110Six-Sigma

(0.237) (0.230) (0.237) (0.229) (0.232)1.133***HPWS(0.214)

1.202***HRM(0.204)

0.789***Motivation(0.190)

0.264***Informationsharing (0.082)

0.215 0.391***WO(0.139) (0.137)


0.384*** 0.558***Workplaceparticipation (0.117) (0.115)

-0.160 -0.119Taskintegration (0.122) (0.122)

Log L -451.23*** -444.48*** -515.96*** -438.518** -500.75***N 336 336 374 336 374


The results indicate that high-performance work systems had a significant

positive correlation with the three indicators of training investment – training

expenditures, training coverage and hours of training – as they did with the degree of

training investment in financial statements (see Appendix 2), supporting Hypothesis 1.

Hypothesis 1-2 is also substantiated by the fact that the supporting HRM index,

a sub-dimension of high-performance work system, as well as its elements of

motivation and information sharing were positively and significantly correlated with all

three indices of training investment.

High-involvement work organization index, the core of high-performance work

systems, is in a significant positive relationship, except with the hours of training per

worker and training expenditures1 in financial statements. When supporting HRM is

omitted from the analysis, it shows a significant positive correlation with the per capita

hours of training and training expenditures1 in financial statements as well, generally

supporting Hypothesis 1-1. As seen in the correlation table, workplace participation, an

indicator of workers participation in developmental activities at workplace, displayed

the strongest relationship with training investment out of all other elements of high-

involvement work organization.

The coefficients of other factors were mostly similar in results to those of

correlation analysis. However, the regression coefficient of competition intensification

and competitive strategy often showed negative, though statistically not significant. As

mentioned in the correlation analysis, since the variable of competition intensification

and competitive strategy could impact training investment through the work system, a

method such as two-stage regression analysis could be employed to deduce the effect.

5. Summary

Based on the hypothesis that high-performance work system is a demand factor

for workforce skills in companies, this study analyzed the effects of high-performance

work practices on firm training. The results showed that high-performance work system

and its elements of supporting HRM and high-involvement work organization increased

corporate investment in training. This finding is similar to other studies that looked into

the relationship between high-performance work systems and firm training.

Corporate investment in training is one dimension of supporting HRM, which

is the sub-dimension of high-performance work system. Therefore it would be more

reasonable to interpret the link between high-performance work system and training

investment as a co-variance, rather than a causal one.

In should also be noted that high-performance work system requires not only

task-related skills but also social skills such as problem solving or self-governance,

suggesting that workers need to be trained in these skills as well.

Although this study is about the demand aspect of firm training, it also makes

suggestions on the supply of skills in general. The overall supply of high-quality skills

required by high-performance work systems would expedite the introduction of those

systems. Given the nature of skills as quasi-public goods, high-performance work

systems do not have enough enticements for individual companies because they require

investments in workforce skills. Therefore, individual companies would find more

incentives to adopt high-performance work systems when high-quality skills are

provided by the society.

The policy suggestions derived from the analysis could be summarized into the

following three points.

First, government policies promoting firm training should be driven in

conjunction with the demand factors of work system. The introduction of high-

performance work systems could prove to be more effective in promoting firm training

than any incentive or regulation. The contents of employee training should also cover

not only task-related skills, but also social skills or self-governing skills. In step with

changes in the nature of work brought about by the replacement of machinery with

electronic equipment, basic education on electronics should be strengthened as well.

Second, tripartite partnership among workers, employers and the government is

needed to supply skills to the society. The supply of high-quality skills would indeed

expedite the introduction of high-performance work systems, which in turn would

increase corporate investment in training. This would ultimately lead to a virtuous cycle

of skill formation in the society. Besides enhancing corporate competitiveness, high-

performance work systems would also raise the quality of working life, resulting in a

win-win game for both the labor and the management. Thus, tripartite dialogues at the

national, regional or by-industry level on high-performance work systems and

workforce skills should be further encouraged.

Third, data on training need to be further developed. As pointed out in this

study, data obtained from the employment insurance database, corporate financial

statements and survey questionnaires all have limitations as measurements of corporate

investment on training.

Although this study attempted to identify the demand factors of firm training, it

came short of analyzing them completely. The demand factor of high-performance work

systems, which in turn is a demand factor of firm training, may be a corporate

competitive strategy in the product market. Researches on the sequential relationship

among corporate competition strategies, high-performance work systems and firm

training are needed in the future.

Regarding the assumption that social supply of skills would accelerate the

introduction of high-performance work systems, researches into the relationship

between training and high-performance work systems should be conducted on the

corporate level. For instance, companies that make heavy investment in employee

training may show higher high-performance work systems adoption rate. Such likely

hypothesis should be verified by using longitudinal data at corporate level.

※ The author welcomes any use of this material provided the source is acknowledged. Nothing

written here is to be construed as necessarily reflecting the views of the Korea Labor Institute.

<References>

In KoreanKim, Dongbae (2000), Factors Influencing the Formal Training of Production Workers,

Management Research, 7:2, 119-134.Kim, Dongbae & Roh Yong-jin. (2002), Determinants of Firm Training, paper presented at the

Korea Corporation Management Association Winter Conference.Kim, Hyung-ki. (1999), Characteristics and Reform Directions of Korean Manufacturers’

Production System, KLI.Kim, Hoon & Kim, Dongbae (2001), A Study on the Introduction of High-involvement Work

Systems, KLI.Ryu, Jang-soo (1997), Korean Manufacturers’ Investment in Training and its Determinants,

Economics Research, 45:4, 227-249.Jeong, Jin-ho (1999), A Study on the Job Mobility in Korea, Seoul National University,

Department of Economics, doctoral dissertation.Joo, Nak-sun (1999), Antecedents and Consequences of Up-skilling Strategy, Seoul National

University, College of Business Administration, doctoral dissertation.

In EnglishAbo, T.(1994), Hybrid Factory: The Japanese Production System in the United States. New

York: Oxford Univ. Press.Appelbaum, E., T. Bailey, P. Berg, A. L. Kalleberg(2000), Manufacturing Advantage: Why

High-Performance Work Systems Pay off. Ithaca, NY.: Cornel Univ. Press.Blauner, R. (1964), Alienation and Freedom, Chicago: University of Chicago Press.Braveman, H.(1974), Labor and Monopoly Capital, New York: Monthly Review Press,.Bright, J. R.(1958), Does Automation Raise Skill Requirement?, Harvard Business Review, 36:4,

85-98.Brown, C., Reich, M., & Stern, D.(1993), Becoming a High-Performance Work Organization:

The Role of Security, Employee Involvement, and Training, International Journal ofHuman Resource Management, 4:2, 247-275.

Cappelli, P., Bassi, L., Katz, H., Knoke, D., Osterman, P., Useem, M.(1997), Change at Work,New York: Oxford Univ. Press.

Cappelli, P., D. Newmark(1999), Do High Performance Work Practices Improve Establish-Level Outcomes? NBER working paper, No. 7374.

Cappelli, P., D. Newmark(2001), External Job Churning and Internal Job Flexibility, NBERWorking Paper. No. 8111.

Cotton, J.L.(1993), Employee Involvement : Methods for Improving Performance and Work

Attitudes, London : Sage.Delaney, J. T.(1996), Unions, Human Resource Innovations, and Organizational Outcomes,

Advances in Industrial and Labor Relations, Vol.7, 207-245.Doeringer, P. B., C. Evans-Klock, D. G. Terkla(1998), Hybrids or Hodgepodges?

Workplace Practices of Japanese and Domestic Startups in the United States, Industrialand Labor Relations Review, 51:2, 171-186.

Felstead, A., F. Green(1996), Training Implications of Regulation Compliance and BusinessCycles, in A. L. Booth, D. J. Snower(eds.), Acquiring Skills: Market Failures, theirSymptoms and Policy Responses. Cambridge Univ. Press, 257-283.

Form, W.(1987), On the Degradation of Skills, Annual Review of Sociology, Vol. 13, pp.29-47.Form, W., R. L. Kaufman, T. L. Parcel, & M. Wallace(1994), The Impact of Technology on

Work Organization and Work Outcome, in G. Farkas, P. England(eds.), Industries,Firms, and Jobs. New York: Aldne de Gruyter, pp. 303-328.

Frazis, H. M. Gittleman, M. Joyce(2000), Correlates of Training: An Analysis Using bothEmployer and Employee Characterics, Industrial and Labor Relations Review, 53:3,443-462.

Freeman, R. B., Medoff, J. L.(1984), What Do Unions Do?, New York: Basic Books.Gardner, T.M., L. M.Moynihan, H.J.Park, P.M.Wright(2001), Begining to Unlock the Black

Box in the HR Firm Performance Relationship: The Impact of HR Practices onEmployee Attitude and Employee Outcomes, Center for Advanced Human ResourceStudies, Cornell Univesity, Working Paper 01-12.

Ichniowski, C., Shaw, K., Prennushi, G.(1995), The Effects of Human Resource ManagementPractices on Productivity, NBER Working Paper No. 5333.

Keenoy, T.(1995), Japanese Managerial Practices in the Maleysian Electronics Industry: TwoCase Studies, Journal of Management Studies, 32:6, 747-766.

Kelley, M. R (1990), New Process Technology, Job Design, and Work Organization: AContingency Model, American Sociological Review, Vol. 55, (April), pp. 191-208.

Knoke, D., A. Kalleberg(1994), Job Training in U.S. Organizations, American SociologicalReview, Vol. 59(Aug.), 537-546

Lawler, E.E.Ⅲ, Mohrman, S.A., Ledford, G.E,Jr.(1998), Strategies for High PerformanceOrganizations, San Francisco: Jossey-Bass Publisher.

Lynch, L. M.(1998), A Needs Analysis of Training Data: What De We Want, What Do WeHave, Can We Ever Get It?, in Haltiwanger, J., E. M. Nanser, & R. Topel(eds.), LaborStatistics Measurement Issues. Chicago: Univ. of Chicago Press. 405-430.

Lynch, L. M., S. E. Black(1998), Beyond the Incidence of Employer-Provided Training,Industrial and Labor Relations Review, 52:1, 64-81.

MacDuffie, J. P., T. A. Kochan(1995), Do U.S. Firms Invest Less in Human Resources?

Training in the World Auto Industry, Industrial Relations, 34:2, 147-168MacDuffie, J.P.(1995), Human Resource bundles and Manufacturing Performance:

Organizational Logic and Flexible Production System in the World Auto Industry,Industrial and Labor Relations Review, 48:2, 197-221.

Osterman, P.(1994), How Common is Workplace Transformation and Who Adopts it?,Industrial and Labor Relations Review, 47:2, 173-188.

Osterman, P.(1995), Skill, Training, and Work Organization in American Establishments,Industrial Relations, 34:2, 125-146.

Parks, S.(1995), Improving Workplace Performance: Historical and Theoretical Context,Monthly Labor Review, 118:5, 18-28.

Pfeffer, J(1994) Competitive Advantage through People: Unleashing the Power of theWorkforce, Harvard Business School Press, MA

Pil, R. K., MacDuffie, J. P.(1996), The Adoption of High-Involvement Work Practices,Industrial Relations, 35:3, pp. 423-455.

Scott, W.R., J. W. Meyer(1994), The Rise of Training Programs in Firms and Agencies, in Scott,W. R., J. W. Meyer(eds.), Institutional Environments and Organizations, London: Sage,228-254.

Smith, A., P.J. Dowling(2001), Analyzing Firm Training: Five Propositions for Future Research,Human Resource Development Quarterly, 12:2, 147-167.

Spenner, K. I.(1995), Technological Change, Skill Requirements, and Education: The Case forUncertainty, in D.B. Bills(ed.), The New Modern Times. Albany: Univ. of New YorkPress, pp. 81-137.

Wagar, T. H.(1997), Determinants of formal Training in Large Canadian Organization,International Journal of Training and Development, 1:2, 82-90.

Whitfield, K.(2000), High-Performance Workplaces, Training, and the Distribution of Skills,Industrial Relations, 39:1, 1-25

Wright, P. W., W.R. Boswell(2002), Desegregating HRM: A Review and Synthesis of Microand Macro Human Resource Management Research, Center for Advanced Human ResourceStudies, Cornell University, Working Paper 02-11.

<Appendix 2> High-performance work system and Training Expenditures(OLS)

M1 M2 M3 M4 M5

3.645*** 3.719*** 3.421*** 3.044*** 3.006***Constants

(0.523) (0.511) (0.531) (0.529) (0.524)0.002 0.001 0.001 0.001 0.000

Age(0.009) (0.009) (0.008) (0.009) (0.008)0.090 0.040 0.155 0.033 0.152

Size(0.098) (0.098) (0.099) (0.100) (0.098)0.008 0.102 0.044 0.098 0.021

Heavy industry(0.228) (0.225) (0.231) (0.226) (0.228)0.110 0.174 -0.045 0.201 0.034

Listed company(0.224) (0.220) (0.226) (0.220) (0.223)0.377 0.503** 0.352 0.447* 0.299

Trade Union(0.246) (0.244) (0.251) (0.247) (0.249)0.123 0.105 0.200 0.107 0.218*Competitive

Pressure (0.122) (0.119) (0.123) (0.121) (0.122)-0.059 -0.055 -0.073 -0.044 -0.055

High-road strategy(0.130) (0.127) (0.130) (0.127) (0.128)0.026 0.117 0.126 0.042 0.006

Six-Sigma(0.267) (0.263) (0.278) (0.267) (0.276)

1.144***HPWS

(0.234)1.136***HRM(0.239)

0.803***Motivation(0.236)0.224**Information

sharing (0.110)0.223 0.424**WO

(0.173) (0.165)0.111 0.122Work group

Autonomy (0.098) (0.099)0.243 0.483***Workplace

Participation (0.153) (0.143)-0.083 -0.095Task

integration (0.149) (0.152)F 5.27*** 5.71*** 3.08*** 4.62*** 3.26***

Adj. R2 .224 .262 .116 .262 .148N 134 134 144 134 144


High Performance Work Practices and Firm Training

Documents