19
Job control and occupational health: the moderating role of national R&D activity KEVIN DANIELS 1 * , OLGA TREGASKIS 2 AND JONATHAN S. SEATON 1 1 Business School, Loughborough University, U.K. 2 Department of Human Resource Management, Leicester Business School, De Montfort University, U.K. Summary Workers (n ¼ 17 275) from 14 European Union (EU) member states provided data on job control, job dissatisfaction, perceived risk of occupational stress, and absence. For each state, level of research and development (R&D) activity was assessed. Associations between individual levels of control and occupational health were stronger where national R&D activity was higher. The moderation occurred for individuals’ levels of control in relation to job dissatisfaction, perceived risk of occupational stress, and absence. The findings with job dissatisfaction and absence were replicated in a sample of workers from 10 Eastern European former Communist countries (n ¼ 7926). Copyright # 2007 John Wiley & Sons, Ltd. Introduction Job control, as a characteristic of an individual’s job, occupies a prominent position in our understanding of how the psychosocial work environment influences occupational health and performance. Research and theorizing concerning job control has originated in many cultural contexts (e.g., de Jonge & Dormann, 2002; Hackman & Oldham, 1980; Karasek, 1979; Warr, 1987), and organizational interventions that include enhanced job control appear to have been widely adopted internationally (Clegg et al., 2002). Much research has tended to focus on how the relationship between control and health is shaped by the work environment (e.g., de Lange, Taris, Kompier, Houtman, & Bongers, 2003) or individual differences (e.g., Bond & Bunce, 2003; de Rijk, Le Blanc, Schaufeli, & de Jonge, 1998; Parker & Sprigg, 1999). In spite of the international breadth of research on job control and health, there has been little consideration of the role of national variables (Parker, Wall, & Cordery, 2001b). In this paper, we examine how aspects of national industrial and economic structures moderate the relationship between individuals’ job control and their occupational health. In particular, we focus on national research and development (R&D) activity. Using the concepts of national business systems (Hall & Soskice, 2001; Whitley, 2000) and operational uncertainty (Wall, Cordery, & Clegg, 2002), we derive hypotheses concerning how national R&D activity may moderate the job control-occupational health relationship. By drawing upon theoretical perspectives from different levels, and examining Journal of Organizational Behaviour J. Organiz. Behav. 28, 1–19 (2007) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/job.390 *Correspondence to: K. Daniels, Business School, Loughborough University, Leicestershire LE11 3TU, U.K. E-mail: [email protected] Copyright # 2007 John Wiley & Sons, Ltd. Received 14 March 2005 Revised 27 February 2006 Accepted 22 March 2006

Job control and occupational health: the moderating role of national R&D activity

Embed Size (px)

Citation preview

Page 1: Job control and occupational health: the moderating role of national R&D activity

Job control and occupational health: themoderating role of national R&D activity

KEVIN DANIELS1*, OLGA TREGASKIS2 AND JONATHAN S. SEATON1

1Business School, Loughborough University, U.K.2Department of Human Resource Management, Leicester Business School, De Montfort University,U.K.

Summary Workers (n¼ 17 275) from 14 European Union (EU) member states provided data on jobcontrol, job dissatisfaction, perceived risk of occupational stress, and absence. For each state,level of research and development (R&D) activity was assessed. Associations betweenindividual levels of control and occupational health were stronger where national R&Dactivity was higher. The moderation occurred for individuals’ levels of control in relation tojob dissatisfaction, perceived risk of occupational stress, and absence. The findings with jobdissatisfaction and absence were replicated in a sample of workers from 10 Eastern Europeanformer Communist countries (n¼ 7926). Copyright # 2007 John Wiley & Sons, Ltd.

Introduction

Job control, as a characteristic of an individual’s job, occupies a prominent position in our

understanding of how the psychosocial work environment influences occupational health and

performance. Research and theorizing concerning job control has originated in many cultural contexts

(e.g., de Jonge & Dormann, 2002; Hackman & Oldham, 1980; Karasek, 1979; Warr, 1987), and

organizational interventions that include enhanced job control appear to have been widely adopted

internationally (Clegg et al., 2002). Much research has tended to focus on how the relationship between

control and health is shaped by the work environment (e.g., de Lange, Taris, Kompier, Houtman, &

Bongers, 2003) or individual differences (e.g., Bond & Bunce, 2003; de Rijk, Le Blanc, Schaufeli, &

de Jonge, 1998; Parker & Sprigg, 1999). In spite of the international breadth of research on job control

and health, there has been little consideration of the role of national variables (Parker, Wall, & Cordery,

2001b).

In this paper, we examine how aspects of national industrial and economic structures moderate the

relationship between individuals’ job control and their occupational health. In particular, we focus on

national research and development (R&D) activity. Using the concepts of national business systems

(Hall & Soskice, 2001; Whitley, 2000) and operational uncertainty (Wall, Cordery, & Clegg, 2002), we

derive hypotheses concerning how national R&D activity may moderate the job control-occupational

health relationship. By drawing upon theoretical perspectives from different levels, and examining

Journal of Organizational Behaviour

J. Organiz. Behav. 28, 1–19 (2007)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/job.390

*Correspondence to: K. Daniels, Business School, Loughborough University, Leicestershire LE11 3TU, U.K.E-mail: [email protected]

Copyright # 2007 John Wiley & Sons, Ltd.

Received 14 March 2005Revised 27 February 2006Accepted 22 March 2006

Page 2: Job control and occupational health: the moderating role of national R&D activity

interactions between national and individual-level phenomena, we hope to show the importance of

national industrial and economic contexts to understanding one important area of organizational

behavior.

Control and occupational health

Job control can be defined as the extent of authority to make decisions concerning the job (Karasek,

1979), and focuses attention on how jobs allow individuals to manage and execute their primary job

tasks (Wall, Wood, & Leach, 2004). For example, job control relates to workers’ ability to choose work

methods, work schedules, and criteria for performance evaluation (Breaugh, 1985). It is a central

component in some of the most influential theories of job design (Parker et al., 2001b).

Using a variety of methodologies, individuals’ levels of job control have been found to be associated

with, for example, less negative mood (Teuchmann, Totterdell, & Parker, 1999); job satisfaction (e.g.,

Janssen, Peeters, de Jonge, Houkes, & Tummers, 2004); physical health (e.g., Bosma, Marmot,

Hemingway, Nicholson, Brunner, & Stansfeld, 1997); greater skill utilization (Holman &Wall, 2002);

improved safety at work (e.g., Parker, Axtell, & Turner, 2001a); reduced absence (e.g., Bakker,

Demerouti, de Boer, & Schaufeli, 2003); and improved work performance (Wall, Jackson, & Davids,

1992). Several qualitative reviews and meta-analyses indicate the importance of job control to worker

health and well-being (e.g., Cass, Farragher, & Cooper, 2002; de Lange et al., 2003; Parker & Wall,

1998; Spector, 1986; Terry & Jimmieson, 1999). As a component of broader high performance work

practices (Wood, 1999), job control is part of an array of human resource management practices linked

to superior financial and safety performance of organizations (Patterson, West, & Wall, 2004;

Zacharatos, Barling, & Iverson, 2005).

While it is clear that job control is important, the processes by which control is beneficial to the

individual remain somewhat unclear (Parker et al., 2001). For example, the mere perception of control

may be sufficient if there is a universal intrinsic need for control (Ganster, 1989). Execution of control

may be beneficial for psychological health, since it allows individuals to minimize exposure to

unpleasant work events (Miller, 1979). Control might also be beneficial, because it promotes active,

problem-focused coping (Karasek & Theorell, 1990).

Wall et al. (2002) proposed a contingency analysis of the influence of empowerment and

performance. Wall et al. regarded job control as the central element in their treatment of empowerment

(p. 161). Although focused on individuals’ levels of job control, they drew upon a wide range of

literatures, including organizational theory, economics, ergonomics, and human resource management.

They also introduced the notion of operational uncertainty, which is a systemic lack of knowledge

concerning production requirements and processes, including uncertainty about what workers should

do and how it should be done. In these circumstances, the opportunity to make work tasks routine is

lessened because potential problems with product manufacture or service delivery are unpredictable.

Therefore, problems are best solved when they occur, because it is not possible to design work systems

in which problems and their solutions are anticipated.

Wall et al. argued that, under conditions of operational uncertainty, job control promotes

job performance in two ways. First, because problems cannot be anticipated in system design,

job control allows workers the discretion to use their own job knowledge to solve problems.

In contrast, in low-control jobs, workers may have to refer problems to other areas in the

organization in order to be given a solution. Second, the act of solving problems promotes learning,

as individuals learn solutions to new problems. This enhancement of job knowledge then makes

solving similar problems consequently more efficient. There is support for the basic proposition

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

2 K. DANIELS ET AL.

Page 3: Job control and occupational health: the moderating role of national R&D activity

that operational uncertainty enhances control’s influence on performance (Wall, Corbett, Martin,

Clegg, & Jackson, 1990).

Whilst Wall et al. (2002) excluded occupational health from their reasoning, there are reasons to

suspect that aspects of occupational health would have stronger relationships with individuals’ levels

of job control under conditions of uncertainty. For example, the interaction between uncertainty and

job control is related to job satisfaction (Wright & Cordery, 1999). The processes of problem solving

that characterize the underlying processes in Wall et al.’s model also underpin the reasoning in

Karasek and Theorell’s demands-control model (1990). According to Karasek and Theorell, the

enhanced problem solving afforded by control does not merely lead to better performance and

organizational learning, but allows people to cope better with the stressors they encounter in their

work. Uncertainty at work is acknowledged as a work stressor (Warr, 1987). Noting that job control

may promote proactive coping in response to the stressor of uncertainty (Jackson, 1989), Parker,

Turner, and Griffin (2003) argued that job control might be particularly important for health under

conditions of high uncertainty.

Given that the beneficial effects of control for health and related outcomes are likely to be greater

under conditions of operational uncertainty, might national-level variables influence uncertainty? It

seems reasonable to suspect national context is influential because of national variations in

organizational practices, financial markets, and labor markets amongst other things (e.g., Tregaskis &

Brewster, 2006). We consider national R&D activity to be one important source of uncertainty, because

the outcomes of innovation are often uncertain (Moore & Davis, 2004). In the following sections, we

explain how national R&D activity, and associated economic, industrial, and labor market factors,

encourage organizational activities that involve greater innovation and increased elements of

operational uncertainty.

National R&D activity, national business systems, and operational uncertainty

National context is acknowledged as an important, yet under researched, factor in job design (Parker

et al., 2001b). For example, relationships between job design and job satisfaction have been attributed

to national culture (Robert, Probst, Martocchio, Drasgow, & Lawler, 2000). Although culture may be

important, other factors should also be considered. Xie (1996) explained patterns of interactions

between job demands, job control, and occupational levels in a Chinese sample by considering the

cultural, political, and economic environment of China.

Research and development activity is a key factor that promotes economic performance at the

national level (Temple, 1999). Encouraging innovation is one of the economic levers by which past

national economic performance can be translated into future economic performance and national well-

being (Jones, 1995). Innovation, at the individual level, has also been linked to job control (Unsworth &

Parker, 2003). Here, we consider R&D activity to be an important, national-level variable that

moderates the link between job control and occupational health. Because national R&D activity is

linked to increased innovation, and innovation introduces elements of uncertainty (Whitley, 2000), we

expect R&D activity to be associated with greater operational uncertainty.Workers in countries typified

by high R&D activity will, then, be more likely to encounter novel work-related problems during the

course of their work. As noted, workers may find solving such problems easier if they have sufficient

job control (Karasek & Theorell, 1990; Parker et al., 2003; Wall et al., 2002).

Next we explain in greater detail how country differences in R&D activity increase operational

uncertainty. At a general level, national institutional factors shape organizational practices (Kostova,

1999). Hence, national policies can influence organizations to adopt strategies and practices based on

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 3

Page 4: Job control and occupational health: the moderating role of national R&D activity

innovation and higher levels of R&D activity (Porter & Stern, 2002). In turn, this should increase levels

of operational uncertainty. Wall et al. (2002) argued that uncertainty has origins beyond the level of the

individual. The influence of national institutions on R&D activity is one example of how this might be

so. The concept of national business systems is useful in order to explain how this happens (Hall &

Soskice, 2001).

Business systems are the ways in which economic activities are coordinated and controlled, and

include elements related to employment legislation, educational policy, and financial institutions. They

are characterized by inter-organizational power relationships and inter-organizational patterns of

cooperation and competition (Whitley, 1998) that influence how organizations within a country

develop their capabilities and strategies (Chang & Shih, 2004; Langlois & Robertson, 1995; Whitley,

2000). More specifically, these institutional factors either constrain the practices that organizations are

able to adopt or provide opportunities for organizations to exploit (Ferner, 1997). In respect of R&D

activity, this gives rise to nationally distinct innovation systems that vary in terms of the levels of

innovation and technological development of organizations (Lundvall, 1992; Nelson, 1993). For

example, Germany’s strong university-industry relationships, high skill levels, and availability of

capital for technology-intensive ventures have been identified as significant contributors to commercial

developments within the chemical sector ahead of the UK (Arora, Landau, & Rosenberg, 1998).

Elements of national business systems that encourage organizations to engage in higher levels of

R&D activity include skills development systems and labor markets. For example, radical innovation in

product design is dependent on tacit knowledge that is specific to the organization and its processes,

which, in turn, is dependent upon on a stable, committed, and highly skilled workforce. In contrast, an

unstable or flexible and low-skilled workforce is more dependent on explicit, codified, and routinized

knowledge (Cowan & Foray, 1997). Employment and educational policies that encourage development

of high-level scientific and technical skills and low labor turnover are more likely to create situations in

which organizations engage in more radical innovation (Lane & Probert, 2005; Whitley, 2000).

Financial institutions can also be influential. Organizational strategies dependent on high innovation

carry enhanced financial risk. The national patterns of how investors interact with organizations can

then shape whether organizations are more or less likely to adopt high innovation strategies. In

Germany, inter-organizational collaboration is encouraged and long-term partnerships exist between

organizations and banks. These partnerships enable the risks associated with innovation to be shared

and encourage sharing of technology across organizations. In contrast, in the UK, short-term financial

returns and shareholder accountability are critical because of the influence of the financial markets as a

source of investment (Hall, 1993). This encourages inter-firm competition and discourages inter-firm

cooperation and technology exchange (Hall & Soskice, 2001). In turn, this limits the ways in which UK

organizations engage in innovation relative to Germany (Arora et al., 1998).

In sum, national institutional systems influence organizational strategies in a variety of ways (e.g.,

skill development, industrial relations, social policy, employment regulatory systems, financial

institutions) that pervade across industrial sectors. Countries typified by high R&D activity will tend to

have institutional structures that support organizations to engage in innovation (Furman, Porter, &

Stern, 2002), and so workers in such countries will experience more operational uncertainty. Following

from our extrapolation of the Wall et al. (2002) model to occupational health, we expect that:

Hypothesis: The relationship between individual levels of job control and indicators of individual

occupational health will be stronger in countries with greater levels of R&D activity.

The hypothesis indicates that we expect the proposed interaction to manifest itself in the relationship

between individual job control and occupational health. This is consistent with the level of theorizing of

Wall and colleagues, and other job design theories (e.g., Karasek & Theorell, 1990).

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

4 K. DANIELS ET AL.

Page 5: Job control and occupational health: the moderating role of national R&D activity

National-level control variables

Other economic factors could be directly related to occupational health. In our analyses and where data

were available, we controlled for gross domestic product (GDP) per capita, unemployment rates, spend

on employment benefit, and inflation rates. Analyses indicate these variables have direct relationships

with life satisfaction (Di Tella, MacCulloch, & Oswald, 2003), and therefore may be related to

occupational health. Moreover, GDP per capita is likely to be correlated with R&D activity (Jones,

1995). High unemployment is a direct, country-level index of low job security, and so reflects the level

of uncertainty over job security in a country. Spend on unemployment benefit is an index of resources

available to buffer the effects of job loss. Inflation rates may reflect uncertainty with respect to wage

bargaining, since governments or central banks may seek to inhibit inflation through tolerating higher

unemployment or acting to restrict pay increases (Franzese, 2001). It is also important to control for

industrial sector, because of differences between sectors. Sector was controlled in the analyses through

the analytic techniques used to analyze the data (Bryk & Raudenbush, 1992).

Context

We examined the hypothesis in two contexts, to enhance the generalizability of the results. In the first,

we used data from a large survey of workers in 14Western European, Scandinavian, andMediterranean

states that had been members of the European Union (EU) for several decades at the time of the survey.

Despite many areas of legislation applying to work organizations across the EU, member states have

control of their economies and there exists variation in economic policy, labor market activity, and

working practices between member states (Hall & Soskice, 2001; Mares, 2004). There is also long-

standing cultural variation (Laurent, 1983). Further, there is variation in the institutional structures in

relation to innovation activity (Whitley, 2000). In the second context, we used data from a large survey

of workers in 10 Eastern European states that were formerly members of the Warsaw Pact and became

market economies. As of 2004, they became members of the EU. At the time of the survey, these

countries were in the process of the social and economic reforms that were conditions of membership of

the EU. These ‘transition’ economies also show the diversity of economic policy and working practices

necessary to test the hypothesis. They provide a test of the hypothesis in different economic, political,

and cultural conditions. These economies are not as developed as those of the first sample and the

nature of the institutional structures makes business systems slow to change (Whitley, 1998; Whitley &

Czaban, 1998).

Methods

Design

For both contexts, we used data from large cross-sectional surveys, and cross-referenced the survey

data against economic indicators of R&D activity assessed for each country.

First context: EU Members’ data

Survey sample

Data were taken from the Third European Survey of Working Conditions (ESWC), conducted in 2000

and sponsored by the EU through the European Foundation for Improvement of Living and Working

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 5

Page 6: Job control and occupational health: the moderating role of national R&D activity

Conditions (see Paoli & Merllie, 2001). The ESWC was conducted in 15 states that were members of

the EU in 2000. The countries surveyed were Belgium, Denmark, Germany, Greece, Spain, France,

Republic of Ireland, Italy, Luxembourg, The Netherlands, Austria, Portugal, Finland, Sweden, and

United Kingdom. The purpose of the survey was to provide an overview of working conditions in

Europe, with questions designed specifically for the survey.

Interviewers administered the survey in participants’ own homes and interviews were conducted in

the national language for each country. Questions were standardized across countries and usually had a

closed response format (e.g., answers were made on rating scales). Questions covered physical working

conditions, social working conditions, organization of time and tasks, health and safety, some non-

work activities, demographic data, and descriptive information on participants’ work organizations.

Participants were selected randomly through a two-step procedure. First, households were selected at

random within postal codes. Second, an interviewer approached each household and requested

participation from the person in work whose birthday followed next after the interview date.

Comparisons between the Eurostat Labour Force Survey and the 1995 ESWC indicate that this

sampling method produces samples representative of the wider working population (Paoli & Merllie,

2001, Eurostat is the official statistical office of the EU).

The sample consisted of workers or self-employed (n¼ 21 703, average response rate¼ 58.6

per cent, range 39 per cent {Italy}–76 per cent {Germany}). Excepting Luxembourg, around 1500

interviews were conducted in each state. For the purposes of this paper, Luxembourg was omitted from

the analyses. Luxembourg is a very small country with an unusual economy based on financial services,

daily migration of workers from neighboring EU states to work, and consequently a very high GDP per

capita (STATEC, 2005). This makes comparison with other EU states, which have more balanced

economies, very difficult. Second, since Luxembourg is such a small state, its overall sample size was

roughly half that of all the other states.

For the purposes of these analyses, data were included from people in paid employment at the time of

the survey and who had not been incapacitated from attending work for the whole of the previous year

due to illness or accidents. The final sample comprised 17 275 responses (with listwise deletion of

missing data). Some 56 per cent of the sample were male. The average age was 38.92 years

(SD¼ 11.52). The modal values for tenure with current employers were 10–14 years (15 per cent), and

also 10–14 years (15 per cent) for tenure in current role. Some 17 per cent of the sample were self-

employed, 17 per cent worked part-time, 85 per cent had permanent employment contracts, and 17

per cent worked shifts. Themodal value for working hours was 30–39 hours per week (35 per cent). The

main occupational groups represented were service and retail workers (16 per cent); craft and trades

workers (16 per cent); clerks (15 per cent); technicians and associate professionals (14 per cent); and

professionals (12 per cent). The main industry sectors represented in the analysis were public services

(26 per cent), manufacturing (19 per cent), retail and wholesale (19 per cent), and financial and

professional services (11 per cent).

Job control

Job control was assessed by six items. The items assessed whether participants were able to influence

their order of tasks, methods of work, rate of work, working hours, breaks, and holidays. Developed

specifically for the survey by an expert panel (see Paoli & Merllie, 2001), these items map onto the

major elements of job control, concerned with influence over the nature, scheduling, and objectives for

work (Breaugh, 1985). For all items, responses were ‘yes’ (coded as 1), ‘no’ (coded as 0), or ‘don’t

know’ (coded as missing). Principal components analysis indicated the presence of one large

component that could explain these six items. This component accounted for over 35 per cent of the

variance, with an eigenvalue of 2.75. The next largest component had an eigenvalue of 1.09, and

accounted for 18 per cent of the variance. The reliability of this index was a¼ 0.76.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

6 K. DANIELS ET AL.

Page 7: Job control and occupational health: the moderating role of national R&D activity

Dependent variables

We tested the hypothesis in relation to three variables: job dissatisfaction, perceived risk from

occupational stress, and absence due to occupational ill-health.

Job dissatisfaction was assessed by a single item, rated on a four-point scale (1¼ ‘very satisfied,’

4¼ ‘not at all satisfied’). Single-item measures of satisfaction are considered reliable and valid

(Wanous, Reichers, & Hudy, 1997). Job dissatisfaction was used as an indicator of psychological ill-

health, since it is central to poor psychological well-being in the workplace (Warr, 1990).

Perceived risk of occupational stress was also used as an index of psychological ill-health. Perceived

risk reflects a consideration that work is a cause of psychological or psychosomatic ill-health.

Perceived risk of occupational stress was assessed by seven items found by principal components

analysis of the ESWC data to form a coherent scale (Daniels, 2004). Participants were asked whether

they believed work affected their health in a number of ways (‘stress,’ ‘anxiety,’ ‘irritability,’ ‘sleeping

problems,’ ‘stomach ache,’ ‘headaches,’ and ‘overall fatigue’). For each symptom, a response of ‘yes’

was coded as 1, and the number of ‘yes’ responses summed to form the scale (range 0 through 7).

Reliability was acceptable (a¼ 0.73).

Absence due to occupational ill-health may reflect either psychological or physical ill-health serious

enough to require time off work. Participants were asked to indicate the number of days in the previous

12 months they had been absent from work due to health problems caused by work.

EU Members’ R&D activity

Research and development activity was assessed using three indicators: percentage spend on R&D as a

proportion of GDP; number of patents filed with the European Patent Office (EPO) for each country;

and the number of patents granted by the US Patents and Trademarks Office (USPTO) for each country.

Values for each index were provided by figures supplied by Eurostat. An average was taken for each

index for the 5 years preceding the survey (1995, 1996, 1997, 1998, 1999). In five instances (two for

Greece and Portugal, one for Sweden), percentage spend on R&D as a proportion of GDP was missing.

Here the mean of the preceding and succeeding values in the time series for each country was

substituted. Each average was then standardized, and the three averages summed to form an overall

index of national R&D activity. Reliability of this index was calculated on the basis of the three

standardized averages (a¼ 0.99).

National-level control variablesValues for GDP per capita, inflation rates, unemployment rates, and national spend on unemployment

benefit were also collected from figures supplied by Eurostat. GDP per capita was indexed by the

average from 1995 to 1999 of GDP per capita in millions of Euros at 1995 prices (sample statistics for

14 EU countries: mean¼ 18.62, SD¼ 5.80). Because of missing data for three countries in 1995,

inflation was indexed by the average increase in consumer prices between 1996 and 1999 (mean¼ 1.87,

SD¼ 1.03). Unemployment rates were indexed by the average number of unemployed as a percentage

of those active in the labor market for the years 1995 through to 1999 (mean¼ 8.97, SD¼ 3.33).

National spend on unemployment benefit was assessed by the average percentage of GDP spent on

unemployment benefits from 1995 to 1999 (mean¼ 2.26, SD¼ 1.06).

Second context: Eastern European data

In 2001, a slightly modified version of the third ESWC was distributed to 12 countries seeking full

membership of the EU (Paoli & Parent-Thirion, 2003). These were Bulgaria, Estonia, Lithuania,

Latvia, Hungary, Poland, Romania, Slovenia, Slovakia, the Czech Republic, Cyprus, andMalta. A 13th

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 7

Page 8: Job control and occupational health: the moderating role of national R&D activity

country, Turkey was surveyed in 2002. The same sampling and methodology was used for the third

ESWC conducted in EU Member states (n¼ 11 057, average response rate¼ 60.2 per cent, range

23 per cent {Poland}–87 per cent {Bulgaria}). Here, analyses were restricted to the 10 countries that

form the transition economies of formerWarsaw Pact countries in Eastern Europe. This is to control for

supra-national context and timing. Malta and Cyprus are relatively small Mediterranean countries, and

were never associated with the Warsaw Pact. Turkey is politically, geographically, and culturally

different from other European states, being predominantly Muslim with the majority of the country in

Asia. Unlike the other states in the survey, Turkey is not yet a member of the EU. Around 1000

interviews were conducted in each of the countries included in the analyses reported here.

Again, data were included only from those in paid employment at the time of the survey and who

had not been incapacitated from attending work for the whole of the previous year due to illness or

accidents. The final Eastern European sample comprised 7926 responses (with listwise deletion of

missing data). Some 51.8 per cent of the sample were female. The average age was 38.89 years

(SD¼ 11.32). The modal values were 6–9 years (15.6 per cent) for tenure with current employers,

and also 6–9 years (14.3 per cent) for tenure in current role. Some 12.7 per cent of the sample were

self-employed, 7.7 per cent worked part-time, 85.6 per cent had permanent employment contracts,

and 22.5 per cent worked shifts. The modal value for working hours was 40–44 hours per

week (15.6 per cent). The main occupational groups represented were craft and trades workers

(19.2 per cent); technicians and associate professionals (16.6 per cent); service and retail workers

(13.4 per cent); professionals (13.4 per cent); and elementary occupations (e.g., refuse collection,

11.7 per cent). The main industry sectors represented in the analysis were public services

(30 per cent), manufacturing (23 per cent), retail and wholesale (13 per cent), and agriculture,

forestry, and fishing (10 per cent).

All of the substantive variables collected from EU Member states were the same for the Eastern

European sample. However, although the question was the same, absence due to occupational ill-health

was coded as a series of 10 categories from ‘no days’ to ‘100þ days,’ rather than as a continuous variable.

Principal components analysis again indicated the six items assessing job control could be explained by a

single factor (for first factor, eigenvalue¼ 2.54, 43 per cent variance accounted for; next factor,

eigenvalue¼ 1.17, 19 per cent variance accounted for). The reliability of this index was a¼ 0.72. For the

index of perceived risk from occupational stress, reliability was also acceptable (a¼ 0.71).

Eastern European R&D activityBecause the same quantity of data are not available as from EU Member states, R&D activity in the

Eastern European countries was estimated from the average spend on R&D as a percentage of GDP in

the 3 years preceding the survey (i.e., 1998, 1999, 2000). There were no missing values for these data.

For data available from USPTO and EPO for the Eastern European states, R&D spend as a percentage

of GDP seemed to form a coherent scale with these other indicators of R&D activity (average

alpha¼ 0.80). Therefore, spend on R&D as a percentage of GDP seems to be a useful single indicator

of R&D activity in these Eastern European states.

National-level control variables

Only GDP per capita and inflation rates were available for the Eastern European countries in the 3 years

preceding the survey. Therefore, country-level indicators were derived from averages calculated from

the years 1998, 1999, and 2000. GDP per capita was indexed by Purchasing Power Standards (PPS),

which is an index to allow comparison with EUMember states. Avalue of 100 indicates the same GDP

per capita as the EU average for any given year. Inflation rate was again indexed by annual increases in

consumer prices. For the 10 Eastern European countries, the sample mean was 45.02 for GDP per

capita (SD¼ 15.46) and 11.48 for inflation (SD¼ 13.98).

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

8 K. DANIELS ET AL.

Page 9: Job control and occupational health: the moderating role of national R&D activity

Analysis

In both samples, the data structures are hierarchical. Three-level multi-level models were used, in

which individual data were nested within industrial sectors, nested within countries. The HLM-5

package was used to analyze the data (Raudenbush, Bryk, Cheong, & Congdon, 2000).

The hypothesis indicates an interaction between individual levels of control and national levels of

R&D activity. In standard regression, interactions are tested by using the product of the independent

variable and the moderator variable (Cohen & Cohen, 1983). This procedure is inappropriate for

hierarchically structured data when the independent and moderator variables are assessed at different

levels of analysis. Moreover, it is important to ensure that the analyses test interactions that occur

because of differences in control between individuals, rather than interactions that occur because of

mean differences in control between sectors and countries. That is, the analyses should indicate that any

significant interaction reflects an interaction between R&D activity and individual levels of control, not

R&D activity and industry sector average levels of control nor R&D activity and national average

levels of control.

Therefore, the following procedure was used to separate out any moderation of national levels of

R&D activity on individual levels of job control from mean levels of job control at these other levels

(Hofmann & Gavin, 1998; Raudenbush, 1989). First, individual-level control was centered at the mean

for the industrial sector in the country to which the individual belonged. This has the effect of removing

the contribution of industrial sector and country average levels of job control, and therefore allowing

any test of interactive effects to assess the impact of R&D activity on the relationship between

individual levels of control and individual levels of occupational health. Individual job control and

R&D activity were then regressed on the dependent variable. The regression slope for individual

control on the dependent variable was allowed to vary between sectors and countries. The R&D activity

index was regressed on the slope for individual control on the dependent variable. This was a test of the

hypothesis, since this examines the moderating role of R&D activity on individual levels of control.

Because the relationship was hypothesized, one-tailed significance tests were used. The index of R&D

activity was centered at the grand mean for the sample, to help deal with any problems of multi-

collinearity.

For analyses with perceived risk of occupational stress and absence as dependent variables, we used

Poisson regression, since the data are both skewed and consist of counts (i.e., the number of symptoms

attributed to stress, or number of days absent, Snijders & Bosker, 1999). The population-average model

was used in these analyses, since we were interested in whether the hypotheses generalized beyond the

current samples.

In all analyses, the following variables were controlled: gender, age, organizational tenure, tenure in

current role, self-employed versus other employed, part-time versus full-time work, shiftwork versus

no shifts, working hours, and occupational group represented as nine dummy variables. Dummy

variables were left in their raw metric and other variables were centered at the mean for the whole

sample (see Bryk & Raudenbush, 1992; Hofmann & Gavin, 1998, for more information on centering).

All relationships between control variables and dependent variables were constrained to be invariant

across sectors and countries. Because of the nature of multi-level modeling, industrial sector and

country were controlled automatically in the analyses, at the second level and the third level of analysis,

respectively. Additionally, R&D activity as well as the country-level controls of GDP per capita,

inflation rate, unemployment rate, and spend on unemployment benefit were regressed on the

dependent variables.

Supplementary analyses were conducted. In one set, the same models were specified as described

above, but job control was centered at the mean for the whole sample. Although, as noted, this has the

effect of confounding individual levels of control with industry sector and country averages, it also

allows a test of whether R&D activity moderates absolute levels of control at the individual level, as

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 9

Page 10: Job control and occupational health: the moderating role of national R&D activity

well as levels of control relative to sector and country means. In the second set of supplementary

analyses, the same models were specified as for the main analyses, but country-level control variables

were omitted. These analyses were conducted to ensure any evidence that R&D activity moderates the

impact of job control on health was not an artefact of anymulti-collinearity involving R&D activity and

the country-level control variables.

Results

Table 1 shows the sample sizes, means, and standard deviations for the substantive variables for both

samples. Table 2 shows the correlations between the substantive variables for both samples. In both

samples, job control is related to lower dissatisfaction, lower self-reported absence due to occupational

health, and lower perceived risk of occupational stress. Whilst the correlations with absence and

perceived risk of stress are low, the correlations are more substantial with job dissatisfaction in both

samples.

Table 3 shows the results of the multi-level regression analyses for both samples. R&D activity

moderates the slope of individual job control in five out of six instances, for all three dependent

variables in the EU Member state sample, and for job dissatisfaction and occupational ill-health

Table 1. Sample sizes, means, and standard deviations for both samples

nControl

Mean (SD)DissatisfactionMean (SD)

Absencea

Mean (SD)Perceived stressMean (SD)

R&D activitya

Mean (SD)

EU Member states 3.69 (1.93) 1.85 (0.73) 1.91 (11.77) 0.95 (1.42) 0.00 (2.97)Austria 1107 3.69 (1.94) 1.69 (0.68) 2.35 (11.13) 0.49 (.98) �0.02Belgium 1247 3.69 (1.19) 1.79 (0.73) 1.77 (12.54) 0.99 (1.50) 0.73Denmark 1266 4.36 (1.61) 1.50 (0.59) 1.44 (0.97) 0.69 (1.11) 0.57Finland 1169 3.66 (1.78) 1.83 (0.58) 3.05 (12.54) 1.18 (1.52) 3.37France 1173 3.58 (1.98) 2.03 (0.74) 1.75 (11.57) 1.27 (1.69) 1.03Germany 1258 3.31 (1.93) 1.89 (0.65) 1.89 (9.21) 0.77 (1.24) 3.24Greece 1313 3.52 (2.16) 2.27 (0.82) 0.86 (8.67) 1.82 (1.59) �4.15Rep. Ireland 1226 3.62 (2.01) 1.60 (0.68) 0.59 (5.04) 0.37 (0.95) �2.07Italy 1159 3.95 (1.82) 2.01 (0.72) 0.98 (7.66) 1.09 (1.50) �2.06The Netherlands 1352 4.08 (1.67) 1.64 (0.74) 5.03 (20.94) 0.74 (1.19) 1.17Portugal 1271 3.17 (2.14) 2.04 (0.64) 1.40 (12.25) 0.68 (1.11) �4.13Spain 1139 3.25 (1.99) 2.04 (0.71) 1.06 (7.03) 1.09 (1.49) �5.53Sweden 1217 3.97 (1.65) 1.89 (0.74) 1.49 (10.62) 1.32 (1.75) 5.62UK 1278 3.72 (1.95) 1.73 (0.75) 1.47 (10.62) 0.95 (1.43) �0.17

Eastern Europe 3.41 (1.84) 2.16 (0.76) 1.28 (1.09) 1.26 (1.55) 0.00 (2.98)Bulgaria 735 2.76 (1.88) 2.17 (0.91) 1.21 (0.90) 1.19 (1.47) �1.60Czech Repub. 820 3.59 (1.82) 2.07 (0.64) 1.27 (1.04) 1.04 (1.02) 4.15Estonia 813 3.90 (1.69) 2.21 (0.69) 1.26 (1.03) 1.48 (1.68) �0.88Hungary 915 3.60 (1.82) 2.07 (0.73) 1.19 (1.02) 1.04 (1.52) �0.07Latvia 782 3.61 (1.77) 2.16 (0.76) 1.18 (0.84) 1.39 (1.54) �2.95Lithuania 787 3.28 (1.87) 2.15 (0.74) 1.19 (0.93) 1.26 (1.55) �1.76Poland 765 3.14 (1.83) 2.10 (0.75) 1.42 (1.22) 1.27 (1.48) �0.44Romania 725 3.28 (1.89) 2.23 (0.82) 1.18 (0.89) 1.33 (1.52) �2.81Slovakia 857 3.45 (1.84) 2.19 (0.74) 1.34 (1.07) 1.32 (1.64) 0.05Slovenia 727 3.38 (1.76) 2.22 (0.81) 1.60 (1.70) 1.27 (1.60) 6.29

aValues for these variables are derived in different ways for the two samples.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

10 K. DANIELS ET AL.

Page 11: Job control and occupational health: the moderating role of national R&D activity

absence in the Eastern European sample. In each case, the sign for R&D activity is negative. This

indicates that in countries where R&D activity is higher, job control has a stronger association with

health. Figure 1a–e show the form of the interaction for each significant result. In each case, there is a

stronger negative relationship between job control and the indicators of occupational ill-health.

Therefore, in five out of six instances, there is support for the hypothesis. Support for the hypothesis is

not found for perceived risk of occupational stress in the Eastern European sample.

The supplementary analyses support the conclusions from the main analyses. In the analyses where

job control was centered at the overall mean for each sample, the direction and significance of the

effects of R&D activity on the job control-health relationship are the same for each dependent variable

(coefficients of R&D activity on job control slope for EU and Eastern European countries, respectively:

B¼�0.01, p< 0.005, B¼�0.01, p< 0.001, dissatisfaction as dependent variable; B¼�0.03,

p< 0.05, B¼�0.005, p< 0.05, absence as dependent variable; B¼�0.01, p< 0.05, B¼�0.003, ns,

perceived risk of stress as dependent variable). Substantial correlations between the country-level

indicators (Table 1) indicate multi-collinearity is a potential problem in the analyses. However, for each

dependent variable, the direction and significance of the effects of R&D activity on the job control-

health relationship are, again, the same as in the main analyses. The coefficients are the same as those

reported for the supplementary analyses with job control centered at the overall mean for each sample.

Individual levels of job control have main effects on all three indicators of ill-health in the EU

Member state sample, and with dissatisfaction and perceived risk of occupational stress in the Eastern

European sample. In each case, higher levels of control are associated with better health. However,

there remain negative associations between control and some indicators of ill-health in countries with

the lowest R&D activity in each sample (Spain: regression coefficients �0.04, �0.02, dissatisfaction,

and perceived risk of occupational stress, respectively; Latvia, regression coefficients �0.04 for

dissatisfaction), job control has a slight positive association with absence in the Spanish sample (0.03),

and the relationship between absence and job control in Latvia is very close to zero (�0.0002).

Finally, significant variance components indicate that there is substantial variation between EU

Member states in the relationships between, on the one hand, individual levels of control and, on the

other, dissatisfaction and perceived risk of occupational stress. These results indicate other factors,

besides R&D activity, might account for variation between countries in the strength of relationships

between job control and health, at least amongst long-standing members of the EU. Although not a

Table 2. Correlations

1. 2. 3. 4. 5. 6. 7. 8. 9.

Control — �0.20���� �0.05���� �0.05���� 0.16 0.37 �0.22 — —Dissatisfaction �0.24���� — 0.13���� 0.24���� �0.10 �0.27 0.40 — —Absence �0.04���� 0.11���� — 0.17���� 0.75� 0.70� �0.21 — —Perceived stress �0.06���� 0.31���� 0.13���� — �0.41 �0.41 0.09 — —R&D activitya 0.39 �0.36 0.60� 0.02 — 0.95�� �0.29 — —GDP/capitaa 0.60� �0.69�� 0.48 �0.32 0.78�� — �0.40 — —Inflationa �0.17 0.55� �0.45 0.50 �0.73�� �0.74�� — — —Unemployment ratea �0.40 0.55� �0.35 0.54� �0.14 �0.38 0.21 — —Unemployment spenda 0.28 �0.41 0.31 0.00 0.60� 0.64� �0.47 0.19 —

Values below diagonal are for EU Member states, above diagonal for Eastern Europe.aCorrelations are with average values of individual-level variables for each country.�p< 0.05; ��p< 0.01; ����p< 0.0001.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 11

Page 12: Job control and occupational health: the moderating role of national R&D activity

Table

3.Multi-level

regressionanalyses

Regressioncoefficients

EU

dissatisfaction

Absence

aPerceived

stress

aEastern

Europedissatisfaction

Absence

aPerceived

stress

a

Country-level

variables

R&D

activity

0.033

0.099

0.132�

0.027

0.049

�0.012

GDPper

capita

�0.011

0.010

�0.001

�0.005

�0.002

�0.001

Inflationrate

0.073

0.014

0.310�

0.001

�0.001

0.003

Unem

ploymentrate

0.031

�0.052

0.082�

n/a

n/a

n/a

Unem

ploymentspend

�0.087

�0.013

�0.085

n/a

n/a

n/a

Individual-level

variables

Jobcontrol

�0.075��

��0

.150��

�0.066��

��0

.076��

��0

.012

�0.040��

Permanentcontract

�0.022

0.003

0.085��

�0.023

�0.030

�0.053

Self-em

ployed

status

�0.038�

�0.554��

�0.010

0.057

�0.005

0.140

Full-tim

ework

0.034

0.199��

��0

.094��

0.081�

�0.050

0.029

Shiftwork

0.092��

�0.242��

�0.289��

�0.065�

0.073��

0.226��

Male

0.009

�0.145��

��0

.148��

�0.013

�0.021

�0.089��

Workinghours

per

week

0.035��

�0.146��

�0.134��

�0.056

0.016

0.117��

Age

�0.001�

0.019��

��0

.004��

��0

.003��

0.000

�0.005��

Organizationtenure

0.008�

0.092��

�0.031��

�0.003

0.003

0.029��

Jobtenure

�0.002

�0.029��

�0.012�

0.001

0.020��

0.024��

Dummyvariablesforjobb

Arm

edforces

�0.169�

0.213��

�0.160

�0.319��

�0.220

0.407��

Seniorofficials/managers

�0.167��

��0

.269��

�0.257��

��0

.248��

��0

.048

0.400��

Professionals

�0.163��

��0

.849��

�0.382��

��0

.207��

��0

.112��

0.400��

Technicians/associateprofessionals

�0.146��

��0

.322��

�0.271��

��0

.173��

��0

.058

0.281��

Clerks

�0.121��

��0

.396��

�0.034

�0.167��

��0

.108�

�0.028

Service/shopworkers

�0.122��

��0

.371��

�0.106��

�0.159��

��0

.055

0.059

Skilledagricultural/fisheries

0.032

0.389��

��0

.034

0.204��

0.031

0.084

Craft/trades

�0.035

0.086��

�0.085�

0.031

0.054

0.098�

Plant/machineoperators/assem

bly

�0.025

0.106��

�0.166��

��0

.068

�0.031

0.206��

Cross-level

interaction

R&D

activityonjobcontrol

�0.006��

�0.032�

�0.008�

�0.012��

��0

.004�

�0.002

Controlslopevariance

components

Country-before

interaction

0.0004��

�0.011�

0.0014��

�0.001��

�0.0001

0.00003

Country-after

interaction

0.0001�

0.007

0.0008��

0.00004

0.00005

0.00001

Sector-before

interaction

0.0001

0.144��

�0.002��

�0.0002

0.00008

0.005��

Sector-afterinteraction

0.0001

0.143��

�0.002��

�0.0002

0.00008

0.005��

aMulti-level

Poissonregressions.Variance

componentsarederived

forunit-specificmodels,since

variance

components

arenotcalculatedforpopulationaveragemodels.

bElementary

occupationscoded

aszero

ineveryjobdummyvariable.

� p<0.05;��p<0.01;��� p

<0.001.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

12 K. DANIELS ET AL.

Page 13: Job control and occupational health: the moderating role of national R&D activity

2.0

0

2.4

0

+1sd

-1sd

Job

cont

rol

2.1

0

1.7

0

1.9

0

.70

1.4

0

.40

.5

0

.2

0

+1sd

-1sd

Job

cont

rol

a)b)

c)

d)e)

+1sd

-1sd

Job

cont

rol

+1s

d-1

sdJo

b co

ntro

l

+1sd

-1sd

Job

cont

rol

Key

+1 s

d R

&D

act

ivity

-1 s

d R

&D

act

ivity

Figure

1.Form

ofsignificantinteractions:(a)DissatisfactionforEU;(b)Absence

forEU;(c)Perceived

risk

ofstress

forEU;(d)DissatisfactionforEastern

Europe;

(e)Absence

forEastern

Europe

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 13

Page 14: Job control and occupational health: the moderating role of national R&D activity

focus of the analyses, the variance components for industry sector are significant for the job control

slopes on absence and perceived risk of occupational stress in the EU sample. There is also a significant

variance component for the slope of job control on perceived risk of occupational stress in the Eastern

European sample.

Discussion

In the present study, five out of six tests indicate support for the hypothesis that national R&D activity

would moderate the influence of individual levels of job control on indicators of occupational health.

Generally, the results indicate that economic contexts can influence the relationships between

individual job characteristics and occupational health. More specifically, since we expect greater

operational uncertainty in national contexts characterized by high R&D activity, these results extend

Wall et al.’s (2002) model of empowerment in two ways. The results indicate that operational

uncertainty moderates job control to predict health, as well as performance. Second, they indicate that

operational uncertainty embedded in national business systems influences relationships at the

individual-level between job design and health.

National R&D activity and job control at the individual level

The present results indicate that workers benefit more from job control in national business systems that

are characterized by high R&D activity, and by extension, national business systems associated with

higher operational uncertainty. This is consistent with Wall et al.’s explanation that job control allows

individual problem solving and learning, thus improving work performance. The results are also

supportive of Karasek and Theorell’s (1990) explanation for the benefits of job control on health, since

they point to the beneficial effects of problem solving for individuals. At another level of analysis, the

results indicate one process by which macro-economic variables can influence occupational health and

well-being. Studies of macro-economics and happiness tend to examine the linear effects of economic

variables (Di Tella et al., 2003; Frey & Stutzer, 2002; Helliwell & Putnam, 2004). In this study, the

results indicate the viability of models that include interactions between individual-level and macro-

economic variables.

However, irrespective of levels of R&D activity, there remained main effects of control on health in

five of the six relationships tested. There are other possible explanations for these main effects that

might supplement, rather than replace, an explanation of the beneficial effects of job control predicated

on problem solving.

First, the main effects of job control may simply reflect other contingencies, not accounted for here.

The significant variance components between countries for the slopes of job control on job

dissatisfaction and perceived risk of occupational stress in the EU sample indicate other contingencies

do exist at the national level. Further, the analyses indicate that some contingencies may also be

associated with different industry contexts: Table 3 shows significant variance components between

industry sectors for the job control slopes on absence in the EU sample and perceived risk of stress in

both samples. However, because sector was assessed at a very broad level in this study (e.g.,

manufacturing), a fine-grained analysis of specific industries might be more appropriate.

Self-determination and control may be universal needs for all human beings (Ganster, 1989), albeit

needs that might vary in strength between national cultures or even sub-cultures within a country (e.g.,

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

14 K. DANIELS ET AL.

Page 15: Job control and occupational health: the moderating role of national R&D activity

Xie, 1996). In this case, job control meets fundamental human needs, and therefore enhances health and

well-being in all situations, but more strongly in some situations. A stronger version of this hypothesis

is that control is completely culturally bound, but beneficial to individuals in all but the most collectivist

societies. In this explanation, control appears to be beneficial because the countries assessed in both

samples exhibit low (e.g., UK), medium levels (e.g., Germany), and moderately high levels of

collectivism (e.g., Greece), compared to societies with very high levels of collectivism (e.g., Pakistan,

Hofstede, 1980). Similarly, job control might be more beneficial to health in cultures less bound by

hierarchy and rules (Thompson, Ellis, &Wildavsky, 1990). Again, European states vary from very low

(e.g., Denmark) to moderately high on this aspect of national culture (France, Hofstede, 1980). Such

explanations might be supported by the significant variance components at the country level for the

slopes of job control on job dissatisfaction and perceived risk of occupational stress in the EU sample:

the socio-economic environments in the 15 long-standing members of the EU have had opportunities to

sustain divergence from each other (Lane, 1995).

The final explanation for the main effects of job control is not linked to fundamental human needs, or

cultural variation in those needs. There is evidence that job control allows individuals to solve problems

at work, but also that job control allows individuals to avoid work stressors, thus regulating aversive

affective impacts of stressors (Daniels & Harris, 2005). Where individuals attempt to avoid stressors

though executing control over their work, factors at the level of the individual, such as the familiarity of

stressors or extent of habitual avoidance behavior, may be important moderators (Daniels, Harris, &

Briner, 2004). However, even here, contextual explanations might be plausible. Only one relationship

indicated no support for the hypothesis, namely the relationship concerning perceived risk of

occupational stress in the Eastern European sample. Because of a history of collectivist policies and

centralized institutional economic control, individuals, arguably, had little responsibility for production

and service delivery. Therefore, there may have evolved cultural patterns of using job control to avoid

work-related problems in order to minimize any anxiety provoked by the problems. Such cultural

patterns of using control as a coping strategy might also explain why job control appears to confer no

benefits for reducing absence in those countries with, in relative terms, the lowest R&D activity (Spain

and Latvia). It may be that the cultural norm in some countries is to use the greater flexibility over work

scheduling associated with job control (Breaugh, 1985) to avoid work in order to recover more fully

from illness

Conclusions

The results indicate that national-level variables reflecting economic and industrial activity influence

the relationship between variables often examined at the level of the individual. One general

conclusion, then, is that national industrial and economic contexts are important in shaping the

individual experience of work, especially in relation to occupational health. The idea that industrial and

economic factors influence job design has been neglected in much of the literature (Parker et al.,

2001b). However, in formulating the hypothesis, the notion of national business systems was used to

explain how industrial and economic factors at the national level combine to shape organizational

structures and strategies that in turn create contexts in which the relationships between individual-level

phenomena vary systematically. A more specific conclusion, then, is that careful consideration of

national economic and industrial factors and the processes by which they shape individual contexts can

help explain why individual-level phenomena vary systematically between nations.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 15

Page 16: Job control and occupational health: the moderating role of national R&D activity

Acknowledgements

We are grateful to the European Foundation for the Improvement of Living andWorking Conditions for

making available data from the third European Survey of Working Conditions for both 2000 data

collection in EU15 states, and 2001 data collection in the accession and candidate countries. We are

also grateful to Dr. Eusebio Rial-Gonzalez, European Agency for Safety and Health at Work.

Author biographies

Kevin Daniels, PhD, is Professor of Organisational Psychology, Loughborough University. Until

recently an Associate Editor of the Journal of Occupational and Organizational Psychology, he is on

the editorial board of the British Journal of Management. His research interests concern the

relationships between emotion, cognition, and organizational processes.

Olga Tregaskis, PhD, is Senior Research Fellow at Leicester Business School. Her research interests

include knowledge diffusion in international networks, learning and development practices, flexible

working patterns in Europe, and international survey methodology. Previously she worked on the

CRANET surveys of HR practices in Europe, and is currently working on a survey of employment

practices in multinational corporations.

Jonathan Seaton, PhD, is Senior Lecturer in Business Economics at Loughborough University. As

well as an interest in labor market economics and the economics of health and welfare, Jon’s interests

include market diversification of firms, the regulation of privatized utilities, and micro-econometrics

analysis, specifically panel estimation and limited dependent variable techniques.

References

Arora, A., Landau, R., & Rosenberg, N. (1998). Chemicals and long-term economic growth: Insights from thechemical industry. London: Wiley.

Bakker, A. B., Demerouti, E., de Boer, E., & Schaufeli, W. B. (2003). Job demands and job resources as predictorsof absence duration and frequency. Journal of Vocational Behavior, 62, 341–356.

Bond, F. W., & Bunce, D. (2003). The role of acceptance and job control in mental health, job satisfaction, andwork performance. Journal of Applied Psychology, 88, 1057–1067.

Bosma, H., Marmot, M. G., Hemingway, H., Nicholson, A. C., Brunner, E., & Stansfeld, S. A. (1997). Low jobcontrol and risk of coronary heart disease in Whitehall ii (prospective cohort) study. British Medical Journal,314, 558–565.

Breaugh, J. A. (1985). The measurement of work autonomy. Human Relations, 38, 551–570.Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models. Newbury Park, CA: Sage.Cass, M., Farragher, B., & Cooper, C. L. (2002). Health and employment: A review and meta-analysis. SalisburyUK: HSE Books.

Chang, P.-L., & Shih, H.-Y. (2004). The innovation systems of Taiwan and China: A comparative analysis.Technovation, 24, 529–539.

Clegg, C. W., Wall, T. D., Pepper, K., Stride, C. B., Woods, D., Morrison, D., Cordery, J. L., Couchman, P.,Badham, R., Kuenzler, C., Grote, G., Ide, W., Takahashi, M., & Kogi, K. (2002). An international survey of theuse and effectiveness of modern manufacturing practices. Human Factors and Ergonomics in Manufacturing,12, 171–191.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

16 K. DANIELS ET AL.

Page 17: Job control and occupational health: the moderating role of national R&D activity

Cohen, J., & Cohen, P. (1983). Applied multiple regression analysis and correlation analysis for the behavioralsciences (2nd ed.). New York: Erlbaum.

Cowan, R., & Foray, D. (1997). The economics of codification and the diffusion of knowledge. Industrial andCorporate Change, 6, 595–622.

Daniels, K. (2004). Perceived risk from occupational stress: A comparison of 15 European countries.Occupationaland Environmental Medicine, 61, 467–470.

Daniels, K., & Harris, C. (2005). A daily diary study of coping in the context of the job demands-control-supportmodel. Journal of Vocational Behavior, 66, 219–237.

Daniels, K., Harris, C., & Briner, R. B. (2004). Linking work conditions to unpleasant affect: Cognition,categorisation and goals. Journal of Occupational and Organizational Psychology, 77, 343–364.

Di Tella, R., MacCulloch, R. J., & Oswald, A. J. (2003). The macroeconomics of happiness. Review of Economicsand Statistics, 85, 809–827.

Ferner, A. (1997). Country of origin effects and HRM in multinational companies. Human Resource ManagementJournal, 7, 19–37.

Franzese, R. J. (2001). Institutional and sectoral interactions in monetary policy and wage/price-bargaining. In P.Hall, & D. Soskice (Eds.), Varieties of capitalism. Oxford: Oxford University Press.

Frey, B. S., & Stutzer, A. (2002). What can economists learn from happiness research? Journal of EconomicLiterature, 40, 402–435.

Furman, J. L., Porter, M. E., & Stern, S. (2002). The determinants of national innovative capacity. Research Policy,31, 899–933.

Ganster, D. C. (1989). Worker control and well-being: A review of research in the work place. In C. L. Cooper, &I. T. Robertson (Eds.), International review of industrial and organizational psychology (Vol. 4, pp. 3–23).Chichester: Wiley.

Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading, MA: Addison Wesley.Hall, B. H. (1993). The stock market’s valuation of R&D investment during the 1980s. American EconomicReview, 83, 259–264.

Hall, P., & Soskice, D. (2001). Varieties of capitalism. Oxford: Oxford University Press.Helliwell, J. F., & Putnam, R. D. (2004). The social context of well-being. Philosophical Transactions of the RoyalSociety of London B, 359, 1435–1446.

Hofmann, D. A., & Gavin, M. B. (1998). Centring decisions in hierarchical linear models: Implications forresearch in organizations. Journal of Management, 24, 623–641.

Hofstede, G. (1980). Culture’s consequences: International differences in work-related values. Beverly Hills, CA:Sage.

Holman, D. J., & Wall, T. D. (2002). Work characteristics, learning-related outcomes, and strain: A test of competingdirect effects, mediated, and moderated models. Journal of Occupational Health Psychology, 7, 283–301.

Jackson, S. E. (1989). Does job control control job stress? In S. L. Sauter, J. J. Hurrell, Jr., & C. L. Cooper (Eds.),Job control and worker health. Chichester: Wiley.

Janssen, P. P. M., Peeters, M. C. W., de Jonge, J., Houkes, I., & Tummers, G. E. R. (2004). Specific relationshipsbetween job demands, job resources and psychological outcomes and the mediating role of negative work-homeinterference. Journal of Vocational Behavior, 65, 411–429.

Jones, C. (1995). R&D based models of economic growth. Journal of Political Economy, 103, 759–778.Jonge, J. de, & Dormann, C. (2002). The DISC model: Demand induced strain compensation mechanisms in jobstress. In M. F. Dollard, H. R. Winefield, & A. H. Winefield (Eds.), Occupational stress in the serviceprofessions. London: Taylor & Francis.

Karasek, R. A., & Theorell, T. (1990). Healthy work. New York: Basic Books.Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job redesign.Administrative Science Quarterly, 24, 285–308.

Kostova, T. (1999). Transfer of strategic organizational practices: A contextual perspective. Academy of Manage-ment Review, 24, 308–324.

Lane, C. (1995). Industry and society in Europe. Aldershot UK: Elgar.Lane, C., & Probert, J. (2005). Globalisation and labour market segmentation: The impact of global productionnetworks on employment patterns of German and UK clothing firms. In A. Ferner, J. Quintanilla, & C. Sanchez-Runde (Eds.),Multinationals and the construction of transnational practices: Convergence and diversity in theglobal economy. London: Palgrave.

Lange, A. H. de, Taris, T. W., Kompier, M. A. J., Houtman, I. R. D., & Bongers, P. M. (2003). ‘‘The very best of themillennium’’: Longitudinal research and the demand-control-(support) model. Journal of Occupational HealthPsychology, 8, 282–305.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 17

Page 18: Job control and occupational health: the moderating role of national R&D activity

Langlois, R. M., & Robertson, P. L. (1995). Firms, markets and economic change. London: Routledge.Laurent, A. (1983). The cultural diversity of Western conceptions of management. International Studies ofManagement and Organization, 13, 75–96.

Lundvall, B.-A. (1992). National systems of innovation: Towards a theory of innovation and interactive learning.London: Pinter.

Mares, I. (2004). Firms and the welfare state: When, why and how does social policy matter to employers? In P. A.Hall, & D. Soskice (Eds.), Varieties of capitalism. Oxford: Oxford University Press.

Miller, S. M. (1979). Controllability and human stress: Method, evidence, and theory. Behavior Research andTherapy, 17, 287–304.

Moore, G., & Davis, K. (2004). Learning the silicon valley way. In T. Bresnahan, & A. Gambardella (Eds.),Building high-tech clusters: Silicon valley and beyond. Cambridge: Cambridge University Press.

Nelson, R. R. (1993) National innovation systems: A comparative analysis. New York: Oxford University Press.Paoli, P., &Merllie, D. (2001). Third European survey of working conditions. Dublin: European Foundation for theImprovement of Living and Working Conditions.

Paoli, P, Parent-Thirion, A. (2003).Working conditions in the acceding and candidate countries. Dublin: EuropeanFoundation for the Improvement of Living and Working Conditions.

Parker, S. K., & Sprigg, C. A. (1999). Minimizing strain and maximizing learning: The role of job demands, jobcontrol, and proactive personality. Journal of Applied Psychology, 84, 925–939.

Parker, S. K., Axtell, C., & Turner, N. (2001a). Designing a safer work place: Importance of job autonomy,communication quality, and supportive supervisors. Journal of Occupational Health Psychology, 6, 211–228.

Parker, S. K., Turner, N., & Griffin, M. A. (2003). Designing healthy work. In D. A. Hofmann, & L. E. Tetrick(Eds.), Occupational health and safety: A multilevel perspective. San Francisco: Jossey Bass.

Parker, S. K., &Wall, T. D. (1998). Job and work design: Organizing work to promote well-being and effectiveness.London: Sage.

Parker, S. K., Wall, T. D., & Cordery, J. L. (2001b). Future work design research and practice: Towards anelaborated model of work design. Journal of Occupational and Organizational Psychology, 74, 413–440.

Patterson, M., West, M. A., & Wall, T. D. (2004). Integrated manufacturing, empowerment, and companyperformance. Journal of Organizational Behavior, 25, 641–665.

Porter, M. E., & Stern, S. (2002). National innovation capacity. In World economic forum, the global competi-tiveness report 2001–2002. Oxford: Oxford University Press.

Raudenbush, S. (1989). A response to Longford and Plewis. Multilevel Modeling Newsletter, 1, 8–10.Raudenbush, S., Bryk, A., Cheong, Y. F., & Congdon, R. (2000). HLM5: Hierarchical linear and nonlinearmodeling. Lincolnwood, Illinois: Scientific Software International.

Rijk, A. E. de, Le Blanc, P., Schaufeli, W., & Jonge, J. de. (1998). Active coping and need for control as moderatorsof the job-demand-control model: Effects on burnout. Journal of Occupational and Organizational Psychology,71, 1–18.

Robert, C., Probst, T. M., Martocchio, J. J., Drasgow, F., & Lawler, J. L. (2000). Empowerment and continuousimprovement in the United States, Mexico, Poland, and India: Predicting fit on the basis of the dimensions ofpower distance and individualism. Journal of Applied Psychology, 85, 643–658.

Snijders, T., & Bosker, R. (1999).Multilevel analysis: An introduction to basic and advanced multilevel modeling.Thousand Oaks: Sage.

Spector, P. E. (1986). Perceived control by employees: A meta-analysis of studies concerning autonomy andparticipation at work. Human Relations, 39, 1005–1016.

STATEC (2005). Retrieved April 5th, 2005, 18, 2000, from STATE website, the Luxembourg Central Services forStatistics and Economic Studies www.statec.public.lu.

Temple, J. (1999). The new growth evidence. Journal of Economic Literature, 37, 112–156.Terry, D. J., & Jimmieson, N. L. (1999). Work control and employee well-being: A decade review. In C. L. Cooper,& I. T. Robertson (Eds.), International review of industrial and organizational psychology (Vol. 14, pp. 95–148).Chichester: Wiley.

Teuchmann, K., Totterdell, P., & Parker, S. K. (1999). Rushed, unhappy, and drained: An experience samplingstudy of relations between time pressure, perceived control, mood, and emotional exhaustion in a group ofaccountants. Journal of Occupational Health Psychology, 4, 37–54.

Thompson, M., Ellis, R., & Wildavsky, A. (1990). Cultural theory. Boulder: Westview.Tregaskis, O., & Brewster, C. (2006). Converging or diverging? A comparative analysis of trends in contingentemployment practice in Europe over a decade. Journal of International Business Studies, 37, 111–126.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

18 K. DANIELS ET AL.

Page 19: Job control and occupational health: the moderating role of national R&D activity

Unsworth, K. L., & Parker, S. K. (2003). Proactivity and innovation: Promoting a new workforce for a newworkplace. In D. Holman, T. D.Wall, C.W. Clegg, P. Sparrow, &A. Howard (Eds.), The newworkplace: A guideto the human impact of modern working practices. Chichester: Wiley.

Wall, T. D., Corbett, J. M., Martin, R., Clegg, C. W., & Jackson, P. R. (1990). Advanced manufacturing technology,work design, and performance: A change study. Journal of Applied Psychology, 75, 691–697.

Wall, T. D., Cordery, J. L., & Clegg, C. W. (2002). Empowerment, performance, and operartional uncertainty: Atheoretical integration. Applied Psychology: An International Review, 51, 146–169.

Wall, T. D., Jackson, P. R., & Davids, K., (1992). Operator work design and robotic system performance: Aserendipitous field study. Journal of Applied Psychology, 77, 353–362.

Wall, T. D., Wood, S. J., & Leach, D. J. (2004). Empowerment and performance. In C. L. Cooper, & I. T. Robertson(Eds.), International review of industrial and organizational psychology (Vol. 19, pp. 1–46). Chichester: Wiley.

Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job satisfaction: How good are single item measures.Journal of Applied Psychology, 82, 247–252.

Warr, P. B. (1987). Work, unemployment and mental health. Oxford: Oxford University Press.Warr, P. (1990). The measurement of well-being and other aspects of mental health. Journal of OccupationalPsychology, 63, 193–210.

Whitley, R. (1998). Internationalisation and varieties of capitalism: The limited effects of cross-nationalcoordination of economic activities on the nature of business systems. Review of International PoliticalEconomy, 5, 445–481.

Whitley, R. (2000). The institutional structuring of innovation strategies: Business systems, firm types, and patternsof technical change in different market economies. Organization Studies, 21, 855–886.

Whitley, R., & Czaban, L. (1998). Institutional transformation and enterprise change in an emergent capitalisteconomy: The case of Hungary. Organization Studies, 19, 259–280.

Wood, S. (1999). Human resource management and performance. International Journal of Management Reviews,1, 367–413.

Wright, B. M., & Cordery, J. L. (1999). Production uncertainty as a contextual moderator of employee reactions tojob design. Journal of Applied Psychology, 84, 456–463.

Xie, J. L. (1996). Karasek’s model in the People’s Republic of China: Effects of job demands, control, andindividual differences. Academy of Management Journal, 39, 1594–1618.

Zacharatos, A., Barling, J., & Iverson., R. D. (2005). High-performance work systems and occupational safety.Journal of Applied Psychology, 90, 77–93.

Copyright # 2007 John Wiley & Sons, Ltd. J. Organiz. Behav. 28, 1–19 (2007)

JOB CONTROL AND R&D ACTIVITY 19