Comparing three alternative types of employment with permanent full-time work: How do employment contract and perceived job conditions relate to health complaints?

Comparing three alternative types of employment withpermanent full-time work: How do employment contractand perceived job conditions relate to health complaints?

CLAUDIA BERNHARD-OETTEL1, MAGNUS SVERKE2 &

HANS DE WITTE3

1National Institute for Working Life, Stockholm, Sweden, 2Department of Psychology, Stockholm

University, Sweden, 3Department of Psychology, Katholieke Universiteit Leuven, Belgium

AbstractPrevious research has found that alternative employment arrangements are associated with bothimpaired and improved well-being. Since such inconsistencies are likely to derive from the type ofemployment contract as well as the characteristics of the job, this paper compares permanent full-timework with forms of alternative employment (permanent part-time, fixed-term and on-call work) inorder to investigate how different employment contracts and perceptions of job conditions relate toindividual well-being. This study contributes to the literature by addressing several questions.Different forms of alternative employment are distinguished and individual background character-istics that might be intertwined with the employment contract are controlled for. Moreover, the scopeof this study extends to the effects of perceived job conditions, and possible interactive effects withtype of employment are tested. Analyses of questionnaire data from 954 Swedish healthcare workersshow that perceptions of the job (job insecurity, job control and demands), but not the type ofemployment contract, predicted health complaints. However, type of employment interacted withperceptions of job insecurity, in that insecurity was associated with impaired well-being amongpermanent full-time workers, while no relationship was found for on-call or core part-time employees.Despite the absence of interactions between employment contract and job demands or job control, itcan be concluded that knowledge about the relationship between alternative employment arrange-ments and the well-being of workers can be enhanced when the combined effects of employmentcontract and job conditions are studied.

Keywords: Alternative employment, temporary work, part-time work, job control, job demands, job

insecurity, health complaints

Introduction

Alternative forms of employment, such as temporary and part-time work, deviate from

traditional ongoing full-time employment concerning the nature and future existence of the

job as well as in the number and distribution of working hours (Reilly, 1998). As a

consequence, they are often assumed to be associated with impaired well-being. However,

the findings of previous research are inconclusive. Some studies, in which temporary or

part-time workers have been found to report more health complaints than those in

Correspondence: Claudia Bernhard-Oettel, National Institute for Working Life, 113 91 Stockholm, Sweden;

e-mail: [email protected]

ISSN 0267-8373 print/ISSN 1464-5335 online # 2005 Taylor & Francis

DOI: 10.1080/02678370500408723

Work & Stress, October-December 2005; 19(4): 301�/318

permanent employment (Benach, Gimeno, & Benavides, 2002; Martens, Nijhuis, Van

Boxtel, & Knottnerus, 1999), support the assumption of detrimental effects of alternative

employment arrangements. In contrast, other studies report better psychological well-being

and fewer health complaints among employees in alternative types of employment (Paoli &

Merllie, 2001; Sverke, Gallagher, & Hellgren, 2000), while yet other studies find no clear

differences (Claes et al., 2002). There may be several explanations to this mixed pattern of

results.

First of all, the majority of studies examining consequences of alternative employment

contracts fail to control for individual and work-related background variables (Virtanen,

M., 2003). Second, research in this area typically does not take into account the

heterogeneity of alternative contracts. Recent research, however, illustrates the necessity

of differentiating between different types of alternative employment arrangements in order

to better understand the consequences for employee well-being (Aronsson, Gustafsson, &

Dallner, 2002; Benavides, Benach, Diez-Roux, & Roman, 2000; Organization for

Economic Co-operation and Development [OECD], 2002). A third possible reason for

the mixed research findings is that the job characteristics of the contract holder are often

overlooked. A growing number of studies suggest that there may be important differences in

job conditions and perceptions between various types of alternative employment, for

instance concerning job insecurity (De Witte & Naswall, 2003; Virtanen, Vahtera,

Kivimaki, & Pentii, 2002), but also concerning job demands and control (Goudswaard &

Andries, 2002). Disadvantageous combinations of job characteristics and job insecurity,

which have been found to be associated with health problems (Strazdins, D’Souza, Lim,

Broom, & Rodgers, 2004), may be more characteristic of some forms of employment than

others (Wikman, Andersson, & Bastin, 1998).

The overarching aim of this paper is to contribute to the understanding of the

relationship between employment contracts and employee well-being. More specifically,

we investigate the direct and interactive effects of employment contract status and perceived

job conditions on individual health complaints, after controlling for individual background

characteristics.

Alternative employment and well-being

Although full-time, open-ended employment contracts still constitute the majority, part-

time and temporary work have become important alternative employment options in most

industrialized countries (OECD, 2002; Sparrow & Cooper, 2003). Alternative employment

can take various forms, with the most common types including fixed-term employment

(e.g. for the duration of a project or as a temporary replacement), on-call employment (e.g.

filling short-term vacancies for a limited number of hours or days) and temporary agency

work (where workers are employed by the agency but temporarily ‘hired out’ to a client

firm) among others (Aronsson, 1999; McLean Parks, Kidder, & Gallagher, 1998; OECD,

2002). Part-time work may be organized on a temporary or*/particularly in Europe*/

permanent basis with benefits comparable to full-time core staff (Krausz, Sagie, &

Bidermann, 2000; OECD, 2002). Thus, permanent part-time workers are yet another

group to be distinguished from ongoing (full-time) and temporary employment (part-time

or full-time).

Despite the fact that many forms of alternative employment exist, they have fre-

quently been studied as a single homogeneous category compared to permanent full-

time employment, which might partly explain the contradictory findings. For example,

some studies report more job dissatisfaction, mental distress and somatic complaints

302 C. Bernhard-Oettel et al.

among temporary workers (Kinnunen & Natti, 1994; Klein Hesselink & Van Vuuren,

1999), while others have reported less mental distress (Letourneux, 1998), fewer somatic

complaints (Sverke et al., 2000), fewer chronic diseases and better self-rated health

among temporary workers compared to permanent staff (Virtanen, Kivimaki, Elovainio,

Vahtera, & Cooper, 2001). Along similar lines, the results are contradictory for part-

time work, with some studies reporting no differences between part-time and full-time

workers (Krausz et al., 2000), while others report higher subjective well-being and job

satisfaction among part-time workers compared to full-time employees (Conway & Briner,

2002).

Recent research suggests that alternative forms of employment may differ along a core�/

periphery continuum with, for example, probationary employment being closer to the core

permanent employees than on-call or seasonal work (Aronsson et al., 2002). With this

perspective, it is also presumed that employment contracts that are more similar to

permanent full-time work will also have similar consequences for well-being. Taking such a

core�/periphery perspective, it may be that different access to fringe benefits (Gustafsson,

Kenjoh, & Wetzels, 2001) and differences in the amount and reliability of income streams

(Beard & Edwards, 1995) influence the well-being of individuals*/particularly in the case

of shorter assignments and high levels of economic dependency on a job (Aronsson,

Dallner, & Lindh, 2000).

Other employment circumstances influencing well-being may be the scheduling of

working time (Martens et al., 1999), prevalence of unfavourable working hours (Paoli,

1996) or duration of service (Sverke et al., 2000). Two recent empirical analyses clearly

demonstrate that different health problems may be more prevalent in certain alternative

forms of employment than others (Aronsson et al., 2002; Benavides et al., 2000). For

example, the study by Aronsson et al. (2002) revealed that discomfort prior to work was

typically reported by substitutes as well as on-call workers; however, while the risk for

stomach complaints was higher among substitutes, back/neck pain was more characteristic

of on-call workers.

Apart from the differences in terms of employment, alternative work arrangements are

spread over different sectors and professions (Aronsson et al., 2002; Davis-Blake & Uzzi,

1993). Hence, alternative forms of employment are found in a variety of jobs that differ

greatly in content and qualification requirements, ranging from manual, low-skilled work to

jobs for the most educated members of the workforce (e.g. a consultant or researcher),

lasting from less than a year to more than 5 years (OECD, 2002). Hence, effects of

employment contract status on health and well-being are difficult to identify if job

conditions are not disentangled from employment conditions (Goudswaard & de Nanteuil,

2000).

Job conditions and well-being

As Goudswaard and Andries (2002) note, the unfavourable job conditions believed

to characterize alternative work arrangements may be reflected in characteristics of

the job, such as limited possibilities to exert influence at work or the terms of employment,

for instance the security of a job and its future prospects. Indeed, empirical studies

often find alternative employment to be associated with poorer job characteristics

(Goudswaard & Andries, 2002; OECD, 2002) and higher job insecurity (Kinnunen &

Natti, 1994; Sverke et al., 2000). However, such a general conclusion may represent

an oversimplification. First, as noted in the previous section, alternative work arrange-

ments are not a single homogeneous group. Indeed, research suggests that poorer

Employment contracts and health complaints 303

job conditions and deviations from conditions in ongoing full-time employment might

be more typical the more peripheral the employment contract (Aronsson et al., 2002;

Wikman et al., 1998). Second, even individuals in comparable employment forms and

positions can perceive job insecurity and job characteristics rather differently, depending on

how a given situation (e.g. employment contract) is interpreted in terms of the future

existence of one’s job or how earlier expectations are met (De Witte & Naswall, 2003;

Virtanen, P. et al., 2002). Hence, not only may different forms of alternative employment

differ in health outcomes, but perceptions of the job may also have an additional influence

on well-being.

The psychological concept of job insecurity refers to subjective concerns about the

continuation of a job (Hartley, Jacobsson, Klandermans, & Van Vuuren, 1991). Feelings of

job insecurity can be considered as a psychosocial risk factor, and the results of a recent

meta-analytic study indicate detrimental consequences for physical and mental health

(Sverke, Hellgren, & Naswall, 2002). However, the majority of previous studies collected

data from permanent employees, for whom such experiences involve a fundamental change

in their perceived position in the labour market (Virtanen, P. et al., 2002). In contrast, for

temporary workers the termination of the job is integrated in their type of employment,

which may lead to less negative reactions. Indeed, two recent studies suggest that the

association between job insecurity and outcomes may depend on the context of the

contractual agreement, as distress increased and job satisfaction diminished significantly

among permanent employees who felt insecure, while temporary workers were unaffected

by perceptions of job insecurity (De Witte & Naswall, 2003; Virtanen, P. et al., 2002).

However, since both these studies differentiated simply between permanent and temporary

workers, it remains unclear whether this conclusion applies to different kinds of alternative

employment. Although job insecurity appears to be related to contractual status, it is not a

perception shared by all individuals in alternative employment arrangements. For instance,

25% of European temporary workers appear to consider their job to be secure (Letourneux,

1998), and part-time workers tend to express fewer feelings of job insecurity than

temporary workers (Sverke et al., 2000). This again illustrates the necessity of differentiat-

ing alternative employment types and of considering subjectivity in perceptions of job

insecurity.

Associations between employment contract and perceived job characteristics in terms of

job control and demands have been investigated in several empirical studies (Aronsson et

al., 2002; Goudswaard & Andries, 2002; Wikman, 2002). Data from the third European

survey on working conditions suggest that high-strain jobs, that is, a combination of high

job demands and low control, are more pronounced among temporary workers than

permanent employees (Goudswaard & Andries, 2002). Similarly, part-time work is often

experienced as low control but, in contrast to temporary work, also low demand, which

results in high monotony and thus passive work (Corral & Isusi, 2004). Again, however,

more can be learnt by a consideration of the heterogeneity of alternative work arrange-

ments. Research contrasting different types of temporary contracts with permanent

employment has found workers in project-based employment to report most influence

and control over their work, whereas substitute employees were in an intermediate position

and on-call workers experienced the fewest possibilities of asserting control (Aronsson et

al., 2002; Wikman, 2002). These results fit well with the findings of Wikman et al. (1998),

who categorized different employment forms using a demand�/control index according to

Karasek’s (1979) model.

304 C. Bernhard-Oettel et al.

The demand�/control model (Karasek, 1979) has been tested in a large number of

studies. Consistent with the model’s assumptions, these studies indicate that psychological

demands tend to be associated with impaired health, increased sick leave and reduced

well-being (Richter et al., 2000), whereas control appears to be related to positive

psychophysiological consequences (Morrison, Payne, & Wall, 2003). An empirical test of

reciprocal relationships (de Jonge et al., 2001) found strong evidence for lagged causal

effects of demands and control on well-being and health. A recent review of high quality

longitudinal studies, however, provides only modest support for the strain hypothesis,

according to which control acts as a moderator buffering the negative influences of demands

(de Lange, Taris, Kompier, Houtman, & Bongers, 2003). Notably, however, the body

of research on job characteristics has been conducted in the context of permanent

employment.

In summary, the combined effects of perceived job characteristics and the features of

various contractual arrangements have received only limited research attention (Aronsson &

Goransson, 1999). Yet there are reasons to assume that the interaction between type of

employment contract and perceived job conditions may have important implications.

Peripheral employment status may interfere with an individual’s capability to cope with

poor job conditions and may impair health considerably. Likewise, demands and control

may yield diverging effects in different contractual arrangements, sometimes even contra-

dicting the theoretically-based assumptions. For some temporary workers, high demands

may represent an opportunity to show one’s skills in order to secure renewed or permanent

employment (Sverke et al., 2000). Due to insufficient levels of information and training

(Aronsson et al., 2002; Corral & Isusi, 2004; Goudswaard & de Nanteuil, 2000), workers in

certain forms of alternative employment may perceive high job control and decision latitude

as threatening and stressful. In other words, there is evidence that type of employment

contract may alter the relationship between job characteristics and employee health and

well-being.

The present study

This paper analyses the direct and interactive effects of employment contract status and

perceived job conditions on employee health complaints in an integrative study. In doing so,

the study extends previous research in several ways. First, adopting a core�/periphery

perspective, we distinguish several alternative forms of employment (part-time work, fixed-

term contracts, and on-call employment) and contrast them with permanent full-time

employment in order to study relationships between employment contract and health

complaints in greater detail. Second, in addition to employment contract status we

investigate perceived job conditions, focusing on job insecurity as well as perceptions of job

control and demands in different employment forms. Beside descriptive associations, we

study the effects of employment contract and subjective perceptions of job conditions in a

combined analysis, thus enabling evaluation of their relative impact on health complaints.

Finally, taking the analysis a step further, we investigate whether job perceptions have the

same effect on individual health complaints in permanent full-time employment compared

to different groups of alternative work arrangements. As individual background character-

istics are often intertwined with employment contract, job conditions and health, we control

in the analyses for age, gender, marital status, professional qualification, organizational

tenure and working hours.

Employment contracts and health complaints 305

Method

Setting and sample

The study is based on questionnaire data collected among the total workforce of two

Swedish hospitals. Questionnaires were mailed by post to the homes of all 2455 employees,

accompanied by a letter explaining the aim of the study, that participation was voluntary,

and assuring that responses were confidential. A total of 1505 participants returned the

questionnaires in the self-addressed envelopes provided, corresponding to a response rate of

61%. Listwise deletion of missing data resulted in an effective sample of 954 individuals

who responded to all the questions that served as the input for the analysis. Of these, 560

(59%) were on permanent full-time contracts, 275 (29%) were permanent part-time

employees, 83 (9%) worked on a temporary basis on fixed-term contracts, and 36 (4%)

were temporary employees working on-call. With respect to their profession, 13% worked

as physicians, 64% as nurses and 23% in administrative positions or as support staff (e.g.

laboratory assistant). The sample consisted of a large majority of women (82%) with a

mean age of 42.5 years. Almost three out of four participants were cohabiting (73%). On

average, the sampled employees worked 36 hours per week and had an organizational

tenure of 11 years.

Measures

Employment contract. Various responses to questions about the employment contract led to the

construction of three categorical dummy variables: permanent part-time (1�/part-time

workers, 0�/other), fixed-term (1�/fixed-term, 0�/other) and on-call (1�/on-call, 0�/

other) with permanent full-time workers as the comparison group (Aiken & West, 1991).

Job conditions. Job insecurity was measured by three items based on Ashford, Lee, and Bobko

(1989). These items refer to a cognitive dimension of job insecurity (e.g. ‘I may be laid off

permanently’) and were assessed on a 5-point Likert scale ranging from 1 (strongly

disagree) to 5 (strongly agree). The a coefficient was .76. Demands were assessed using

three items (e.g. ‘I often have too much to do in my job’) from a scale measuring role

overload (Beehr, Walsh, & Taber, 1976). The a coefficient was .81. Control was assessed

using a 4-item scale (Sverke & Sjoberg, 1994) based on Hackman and Oldham (1975) and

Walsh, Taber, and Beehr (1980) to capture the amount of influence and control over one’s

job (e.g. ‘I have satisfactory influence over decisions concerning my job’). The scale

demonstrated satisfactory internal consistency (a�/ .81). Demands and control were also

assessed on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

Results of confirmatory factor analyses showed that the presupposed three-factor model

provided a satisfactory fit to data [x2 (32)�/378.05; p B/.001; RMSEA�/0.09; AGFI�/

0.90] and clearly outperformed rivalling one- or two-factor models, thus supporting the

existence of three distinct job conditions.

Health complaints. Mental distress was measured using the General Health Questionnaire

(Goldberg, 1979), which consists of 12 questions where respondents indicated how often

they had experienced certain symptoms (0�/never, 3�/always) over the past weeks. The

scale, which is deemed to be a good indicator of non-psychiatric disorders, demonstrated

adequate reliability (a�/.82). House and Rizzo’s (1972) scale was used to assess job-induced

tension . The seven items (e.g. ‘Problems associated with my job have kept me awake at

306 C. Bernhard-Oettel et al.

night’), scored on 5-point Likert scales, yielded a reliability estimate of .84. Again,

confirmatory factor analysis suggested distinct constructs. The results indicated that the

two health complaints measures reflect two distinct factors [x2 (151)�/1830.75; p B/.001;

RMSEA�/0.11; AGFI�/0.80], and the two-factor solution provided a significantly better

fit to data than a one-factor model.

Individual background characteristics. Age and organizational tenure were measured in years.

Weekly working hours were assessed by respondents giving their average working hours per

week. Gender (1�/female, 0�/male) and partner (1�/married or cohabiting, 0�/living

alone) were measured as dummy variables. Two categorical variables were constructed

representing the most frequent occupational groups: physicians (1�/physicians, 0�/other)

and nurses (1�/nurses, 0�/other).

Descriptive statistics (means, standard deviations and intercorrelations) for the variables

are presented in Table I. Continuous variables were screened for normality and the majority

of variables’ distributions were normally shaped. Job insecurity showed a tendency towards

positive skewness and the number of working hours was found to have a positive kurtosis,

indicating a distribution that is slightly too peaked.

Statistical treatment

In order to compare how the different employment contracts differed in the variables under

study, three multivariate analyses of variance (MANOVA) were performed on individual

background characteristics, job conditions and health indicators. Before the use of

MANOVA, Box’s M tests of homogeneity of variance-covariance matrices were conducted.

Significant results for the variables of job conditions and individual background

characteristics indicate that smaller samples produced larger variances and covariances in

these variables. Therefore, Pillai’s trace was used to evaluate multivariate significances in

the MANOVA analyses on job conditions and individual background characteristics. The

MANOVAs were supplemented with univariate F-tests and post-hoc Scheffe tests. For

individual background characteristics assessed as categorical variables, x2 tests were used to

test for differences between employment forms.

For the predictive analysis of the direct and interactive effects of employment contract

types and job conditions on health complaints, a hierarchical regression analysis was used

for each of the dependent variables (mental distress and job-induced tension). In order to

ensure that assumptions for the use of multiple regressions are met, the data was screened

for outliers, multicollinearity and tolerance levels. In each of these regressions, individual

background variables were first entered into the model to control for their influence on the

outcomes. In the second step, the different types of employment contract were included in

the regression equation to find out whether differences in employment contracts explained

variance in the health-related outcomes, while controlling for individual background

characteristics. In step 3, job insecurity, job control and demands were added to the model.

In steps 4, 5 and 6, the interaction terms between employment contracts (part-time, fixed-

term, on-call) and job perceptions (job insecurity, job control, job demands) were included

in the model separately in order to test which of them had any additional significant effect.

The interaction terms were created with the cross-product of the variables. Following the

recommendations by Aiken and West (1991), continuous predictor variables were centred

before the cross-products were calculated in order to avoid artificial multicollinearity.

Employment contracts and health complaints 307

Table I. Correlations, means and standard deviations for all variables, and a reliabilities for scales.

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 M SD a

Individual background characteristics

1 Age 1.00 42.49 10.19 �/

2 Gender

(female)

.01 1.00 0.82 .38 �/

3 Partner .02 .03 1.00 0.73 .44 �/

4 Physician .12* �/.45* .08* 1.00 0.13 .34 �/

5 Nurse �/.18* .26* .00 �/.52* 1.00 0.64 .48 �/

6 Weekly

working hours

.09* �/.20* .00 .42* �/.29* 1.00 36.34 8.64 �/

7 Organizational

tenure

.53* .15* .03 �/.12* .05 .01 1.00 11.38 8.99 �/

Employment contract

8 Part-time .02 .16* .13* �/.20* .19* �/.49* .06* 1.00 0.28 .45 �/

9 Fixed-term �/.27* .00 �/.05 .12* �/.06* .07* �/.27* �/.21* 1.00 0.10 .30 �/

10 On-call �/.12* �/.06* �/.06* �/.07* .02 �/.34* �/.12* �/.13* �/.07* 1.00 0.04 .20 �/

Job conditions

11 Job insecurity �/.07* �/.03 �/.05 �/.03 �/.01 .�/07* �/.03 �/.09* .30* .17* 1.00 1.75 .99 .76

12 Job control .11* .02 .06* �/.04 �/.10* .06* .09* �/.02 �/.09* �/.09* �/.16* 1.00 3.30 .90 .81

13 Job demands �/.04 �/.01 �/.01 .11* �/.08* .12* �/.09* �/.04 �/.02 �/.07* .05 �/.30* 1.00 3.66 1.00 .81

Health complaints

14 Mental distress

(GHQ)

.06* �/.04 �/.11* �/.03 �/.08* -�/.04 .03 �/.01 .05 .05 .23* �/.33* .31* 1.00 1.74 .39 .82

15 Job-induced

tension

�/.06* �/.10* �/.01 .12* �/.10* .15* �/.06* �/.06* .08* �/.08* .19* �/.28* .53* .56* 1.00 2.63 .91 .84

*pB/ .05.

Scale range: 0�/1 (variables 2�/5, 8�/10 [for these variables the mean value symbolizes the proportion scoring 1]), 1�/5 (variables 11�/13, 15), 0�/3 (variable 14), years

(variables 1, 7), hours/week (variable 6).

308

C.

Bern

hard

-Oettel

etal.

Results

Characteristics of different employment contracts

Table II presents mean levels for job conditions and health complaints as well as means/

frequencies for individual background characteristics for full-time, part-time, fixed-term

and on-call workers. It also reports the results of the tests for differences between

employment contracts.

For job conditions, an overall multivariate effect was found (Multivariate F

(df�/9,2850)�/18.58, p B/.001). Univariate F-tests revealed differences in job insecurity as

well as job control and job demands. Scheffe post-hoc tests showed that the level of job

insecurity was higher for fixed-term and on-call workers than for part-time and core full-

time employees. Core full-time workers were also found to have higher levels of control and

demands compared to employees contracted on an on-call basis.

There was also a significant overall difference between employment contracts in health

complaints (Multivariate F (df�/6,1900)�/4.21, p B/.001). Univariate F-tests revealed no

significant differences in mental distress among employment groups, whereas there was a

difference in job-induced tension. The post-hoc tests revealed that job-induced tension was

higher for fixed-term workers than for employees on an on-call contract.

In terms of individual background characteristics, there was a significant multivariate

difference between the employment groups in the MANOVA conducted on age, weekly

working hours and organizational tenure (Multivariate F (df�/9,2850)�/67.23, p B/.001),

and the univariate follow-up tests revealed differences in all three variables. Scheffe post-hoc

tests showed that core full-time workers resembled part-time workers in terms of age, while

the other two groups were on average younger. Core full-time workers reported the highest

number of working hours per week, followed by fixed-term workers, but this number was

significantly reduced for part-time workers and lowest for the group of on-call workers.

Organizational tenure was higher for core full-time and part-time workers compared to

fixed-term and on-call employees. Chi-square tests performed for categorical variables

revealed significant differences in all variables. In terms of the distribution across

occupations, employees on fixed-term contracts were most comparable to full-time workers

while part-time workers resembled on-call contracted workers. The percentage of women

was highest among part-time workers, as was also the proportion of those living with a

partner.

Predicting health complaints

The results of the regression analyses predicting mental distress and job-induced tension are

presented in Table III. In Step 1, all individual background variables except for weekly

working hours and organizational tenure displayed a small but significant association with

mental distress (R2�/.03, pB/ .001). The second step, in which the employment contracts

were added, represented a small increase in explained variance (R2change�/.01; p B/.05), but

none of the contract variables reached significance. In Step 3 all job conditions evidenced

significant effects on mental distress and increased the explained variance to .24 (pB/ .001).

Both job insecurity and job demands were positively related to mental distress, while the

association was negative for job control. In steps 4 to 6, which tested interactions between

employment contract types and each of the perceived job conditions, only the interaction

terms between employment contracts and job insecurity (Step 4) increased the overall

explained variance of the model significantly (R2change�/.02; p B/.001). In comparison with

the core full-time workers, there was a difference in the relationship between job insecurity

Employment contracts and health complaints 309

Table II. Individual background characteristics, job conditions, general health and job induced tension of workers with alternative versus core full time employment.

Employment contract

Variable Full-time [1] Part-time [2] Fixed-term [3] On-call [4] Total Chi square Univariate F Means comparison

Job conditions

Job insecurity 1.60 1.62 2.77 2.64 1.75 �/ 51.24*** 1B/3, 4; 2B/3, 4

Job control 3.39 3.27 3.03 3.00 3.31 �/ 5.65*** 1�/4

Job demands 3.72 3.61 3.56 3.26 3.66 �/ 3.17* 1�/4

Health complaints

Mental distress 0.73 0.74 0.82 0.82 0.74 �/ 1.82 n.s.

Job-induced tension 2.65 2.54 2.86 2.30 2.63 �/ 4.39** 3�/4

Individual background characteristics

Age (mean) 44.2 42.7 34.2 36.5 42.5 �/ 32.14** 1�/3, 4; 2�/3, 4

Gender (% female) 78.6 91.6 79.5 66.7 82.0 27.81*** �/ �/

Partner (% cohabiting) 70.0 82.5 68.7 66.7 73.4 16.87*** �/ �/

Physician% 17.0 2.2 25.3 2.8 12.9 51.02*** �/ �/

Nurse% 58.2 78.9 55.4 72.2 64.5 38.51*** �/ �/

Weekly working hours (mean) 40.4 29.6 37.4 22.2 36.3 �/ 220.73*** 1�/2, 3, 4; 3�/2, 4; 2�/4

Organizational tenure (years) 12.5 12.2 3.7 6.1 11.4 �/ 30.48*** 1�/3, 4; 2�/3, 4

*p 5/.05; **p 5/.01; ***p 5/.001.

310

C.

Bern

hard

-Oettel

etal.

and mental distress only for employees in on-call arrangements, but not for part-time or

fixed-term workers. As can be seen in Figure 1, permanent full-time workers reported

higher mental distress when they perceived high job insecurity (t�/ 5.72, p B/.001), whereas

on-call workers’ mental distress was not dependent on the level of perceived insecurity (t�/

0.34, p �/.05). The results remained the same when the interaction terms were added in a

different order.

0.2

0.4

0.6

0.8

low job insecurity high job insecurity

Full time On-call

0

1

Figure 1. Interaction of perceived job insecurity and type of contract (core full-time and on-call) on mental distress

(scale 0�/3).

Table III. Predicting mental distress and job induced tension: standardized regression coefficients (N�/954).

Mental distress b last step Job-induced tension b last step

Step 1: Individual background characteristics

Age .10** �/.05

Gender (female) �/.07* �/.08**

Partner �/.08** .00

Physician �/.16*** �/.04

Nurse �/.13*** �/.06

Weekly working hours �/.02 .09*

Organizational tenure .04 .04

R2change .03*** .04***

Step 2: Employment contract

Part-time contract .01 .01

Fixed-term contract �/.02 .00

On-call contract .05 �/.02

R2change .01* n.s.

Step 3: Job conditions

Job insecurity .18*** .19***

Job control �/.27*** �/.11**

Job demands .28*** .51***

R2change .20*** .30***

Step 4: Employment x Job Insecurity

Part-time�/Job insecurity �/.05 �/.07*

Fixed-term�/Job insecurity .07 .01

On-call�/Job insecurity �/.11** �/.08*

R2change .02*** .01*

Step 5: Employment�/Job control

R2change n.s. n.s

Step 6: Employment�/Job demands

R2change n.s. n.s.

*p 5/.05; **p 5/.01; ***p 5/.001.

Employment contracts and health complaints 311

Among the individual background variables, only (female) gender and the number of

weekly working hours were positively related to job-induced tension (R2change�/.04; p B/

.001). With the effect of individual background variables held constant, there was no

significant increase in the explained variance when the types of employment contract were

entered in Step 2. All perceived job conditions entered in the third step were significantly

associated with job-induced tension, and the proportion of explained variance increased to

34% (p B/.001). Job demands and job insecurity were positively associated with job-induced

tension, whereas the effect of job control was negative. Again, in steps 4 to 6, only the

interactions between forms of employment contract and job insecurity (Step 4) explained

additional variance in the outcome (R2change�/.01; pB/ .05). Both part-time employees and

on-call workers, but not fixed-term workers, displayed associations between job insecurity

and job-induced tension that differed from full-time permanent workers. As can be seen in

Figure 2, job-induced tension increased significantly for permanent full-time workers when

they perceived high levels of job insecurity (t�/4.66, p B/.001). For core part-time (t�/ 1.52,

p �/.05) and on-call workers (t�/ 0.05, p �/.05), the level of perceived job insecurity did not

affect job-induced tension. These results were unaffected by the order in which interaction

terms were added.

Since there may be a risk of multicollinearity between working hours, organizational

tenure and profession, with employment contract reducing the impact of the predictor

variable, several regressions were run, leaving out one or all of these control factors. In

general, these supplementary analyses produced similar results, with the same pattern of

significant predictors.

Discussion

Previous research on alternative work arrangements and well-being highlights the

importance of considering individual background characteristics of the contract holders

and, in addition, of going beyond a simple distinction between permanent work and a

uniform category of alternative employment contracts. Furthermore, research suggests that

the understanding of the consequences of alternative employment is enhanced by taking

into account other explanatory effects that possibly may be intertwined with the type of

employment contract, among which perceived job conditions are believed to have a major

potential impact. Thus, the overall objective of this study was to investigate the direct and

2.4

2.6

2.8

3

2low job insecurity high job insecurity

Full time

Part time

On-call

2.2

Figure 2. Interaction of perceived job insecurity and type of contract (core full-time, core part-time and on-call) on

job induced tension (scale 1�/5).

312 C. Bernhard-Oettel et al.

interactive effects of employment contract status and job conditions on employee health

complaints once individual background characteristics were controlled for.

In order to broaden the understanding of the characteristics of different types of

employment contract, we initially compared three categories of alternative employment

(part-time, fixed-term and on-call contracts) with permanent employment. Consistent with

previous research, we found some support for a core�/periphery distinction, but only for job

insecurity as well as the distribution of age and organizational tenure (cf. results reported by

Sverke et al., 2000, and OECD, 2002). Fixed-term workers were found to be rather similar

to core full-time employees and clearly different from both part-time and on-call workers.

Part-time workers shared some characteristics with those on full-time contracts (e.g.

relatively positive job perceptions) but were similar to on-call workers in other respects (e.g.

a low number of weekly working hours). As argued in previous research (Aronsson et al.,

2002; Bernhard & Sverke, 2003), these results underline important distinctions among

different types of alternative employment. However, the differences and similarities

obtained in the present data are not clear-cut and challenge the idea of a core�/periphery

perspective.

This complex pattern of similarities and differences between forms of contract may

explain why it remains unclear whether and how health and well-being are related to

differences in employment contracts. When individual background characteristics were held

constant, contract status was unrelated to job-induced tension and only marginally

contributed to the variance explained in mental distress. The fact that these results held

up also when background characteristics that might be related to employment contract

status (i.e. tenure, position or working hours) were excluded from the model calculation

lends further credibility to our findings. Thus, a rather minimal association between

contract and well-being was found (for similar findings, see Krausz et al., 2000), even

though alternative employment types were differentiated. This finding contradicts the

suggested core�/periphery continuum, according to which permanent workers should have

the best and on-call workers the worst conditions of health and well-being (Aronsson et al.,

2002). It is possible that different types of employment contract do not automatically reflect

differences in employment conditions along a core�/periphery continuum. For example,

organizational tenure was initially assumed to be much shorter in contracts localized at the

more peripheral end of the continuum (Sverke et al., 2000), but tenure appeared to be

relatively long in all contracts of this sample. Hence, merely comparing forms of

employment contracts may not reveal much information about differences or similarities

in employee well-being if the specifics of employment terms are not studied.

This study also argued that subjective perceptions of one’s job conditions might be

another possible major explanation of well-being in different employment situations.

Indeed, in agreement with previous research, we found impaired mental health and higher

levels of job-induced tension to be associated with higher perceived job demands and

experiences of lower job control (de Lange et al., 2003) as well as higher levels of job

insecurity (Sverke et al., 2002). All three perceptions increased the proportion of explained

variance in health complaints markedly, and thus our results support the conclusion that

subjective perceptions of job conditions, compared to objective conditions (i.e. contract

status), are stronger predictors of employee health and well-being (for a similar conclusion,

see, for example, Krausz et al., 2000).

This study further examined the potential interaction effects between type of employ-

ment contract and perceived job conditions on health complaints. Type of employment

contract was not found to alter the effects of job demands or job control on employee well-

Employment contracts and health complaints 313

being. In line with earlier studies (De Witte & Naswall, 2003; Virtanen et al., 2002),

however, significant interaction effects were found between types of employment contract

and job insecurity. Fixed-term workers reacted just like core full-time employees, with

impaired well-being when experiencing high job insecurity, while health complaints of

on-call workers and job-induced tension of core part-time as well as on-call workers

were unrelated to job insecurity. Thus, negative effects of job insecurity on health

complaints appear to depend on employment circumstances, and the relevance of

distinguishing different alternative employment contracts becomes obvious. However, the

present results challenge previous findings that job insecurity affects only permanent

workers because it is not part of their employment contract (De Witte & Naswall, 2003;

Virtanen et al., 2002).

Theories of balance of the psychological contract between employer and employee in

terms of efforts and rewards may be added here (Blau, 1964; Siegrist, 1996), suggesting

that if future rewards are promised by the employer or assumed by the employee (e.g. a

prolonged or permanent contract), sacrifices (e.g. high job insecurity) may be acceptable for

employees in the short term, if the balance is re-adjusted afterwards. The individual motives

to take up alternative employment may influence the relevance of job security, since the

effect of job insecurity may, for example, be altered for individuals with lower work

involvement (Hellgren, 2003). In terms of the absence of interaction effects between

employment contract types and perceived control and demands, this exploratory study

cannot draw a firm conclusion on whether this finding shows the non-existence of such an

effect. Since there are no earlier studies for comparison, it may be equally plausible that

relations between perceived job conditions and well-being are altered only for certain

categories of temporary workers (e.g. individuals who have an intermittent relationship with

the organization) or under certain circumstances (e.g. during periods when employers make

decisions about contract renewal).

Limitations

As with other research, there may be limitations to our study that influence its conclusions.

First, although a number of individual background characteristics were controlled for, other

individual as well as work-related variables might be of interest, e.g. household income,

number of children and work involvement. An interesting effect to be addressed in future

research concerns the nature of the relations between background characteristics and

outcomes. Albeit not present in this study, it is plausible that associations between, for

instance, working hours and health outcomes may be U-shaped, such that too few as well as

too many working hours may be associated with health problems. Second, only three types

of alternative work arrangements were studied and compared to permanent full-time work.

Despite the fact that our distinction between different forms of alternative employment

extends the focus of many previous comparative analyses on employment contracts and

well-being, the results presented here do not allow generalization to all forms of alternative

employment. A third limitation is that we studied only certain aspects of job perceptions,

although previous research also shows that factors such as volition (Krausz, Brandwein, &

Fox, 1995), social support (Isaksson & Bellaagh, 2002) and psychological contract

perceptions (De Cuyper, Isaksson, & De Witte, 2005) may be of importance for employees’

health and well-being. Moreover, although the empirical support for interactive effects

between employment contract types and perceived job conditions was limited, it is

conceivable that employment contract interacts with other perceptions of employment,

job conditions or aspects of choice.

314 C. Bernhard-Oettel et al.

Another potential limitation to the present findings concerns sample restrictions. The

current study was restricted to healthcare workers, who may share certain role character-

istics (Sverke et al., 2000) and whose perceptions may be related to circumstances in the

labour market for caring occupations (Isaksson, Bernhard, & Gustafsson, 2003). Expand-

ing the analyses to other organizational circumstances and sector settings would be fruitful,

since several research studies hint at problems of generalization across occupations or

sectors, which may use different alternative contracts with job conditions tailored to their

specific requirements (Cohen, Haberfeld, & Ferber, 1993). Another sample specific that

may have influenced the effects of perceived job conditions and employment circumstances

is the rather high organizational tenure of temporary employees in our study.

Yet another limitation can be seen in the use of cross-sectional data, making causal

inferences more difficult. Although our results, for the most part, are in line with theoretical

propositions and earlier research, it appears to be important to replicate the study

longitudinally, for instance to control for overestimation of the effects of job perceptions.

Further, in this study we chose to limit our analysis to measures of job-related and mental

health complaints. However, a challenging question is to what extent employment

conditions as well as perceptions of job conditions constitute health risks and how they

are associated with diagnosed diseases, which would constitute extra difficulties for

transition into permanent or full-time employment.

Concluding remarks

Despite these limitations, this study enhances scientific understanding of forms of

alternative employment, and adds to previous research in several ways. First, an effort

was made to differentiate between forms of alternative employment. The results show that

alternative employment contracts are associated with important differences, thus indicating

that a basic distinction between core and alternative work forms may be an over-

simplification. However, whether the proposed categorization into a core�/periphery

perspective (Aronsson et al., 2002) is helpful in depicting differences of employment and

job conditions of different contract types remains an open question. There are several

dimensions according to which differences between contracts can emerge, and further

research on other contract types and other branches is needed to evaluate how employment

contracts are related to health.

Second, the results show the benefit of complementing studies on employment contract

effects with the effects of job perceptions. Subjective perceptions of job conditions are

important in understanding consequences for health, but they can furthermore have

combined effects with the form of contract, which may alter their meaning for employees’

well-being in certain employment circumstances. Here, perhaps, practitioners and

researchers can gain more knowledge by analysing precisely what employment conditions

are tied to the contracts under investigation, particularly since variations in forms of

employment are almost infinite and not invariant across sectorial and national boundaries

(McLean Parks et al., 1998).

In conclusion, integrating objective as well as subjective aspects relating to employment

and jobs in analyses of future research seems to be highly necessary in order to shed more

light on the question of how alternative employment arrangements may affect individual

well-being and health.

Employment contracts and health complaints 315

Acknowledgements

The research presented here was financed by the Swedish Council for Working Life and

Social Research and Stockholm County Council. The authors thank Peter Richter for

comments on a previous version of the manuscript.

References

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions . Newbury Park, CA: Sage.

Aronsson, G. (1999). Contingent work and health and safety. Work, Employment & Society, 13 , 439�/459.

Aronsson, G., Dallner, M., & Lindh, T. (2000). Flexibla inkomster och fasta utgifter [Flexible income and fixed

costs]. Arbete och Halsa , 20 , 1�/26.

Aronsson, G., & Goransson, S. (1999). Permanent employment but not in a preferred occupation: Psychological

and medical aspects, research implications. Journal of Organisational Health Psychology, 4 , 152�/163.

Aronsson, G., Gustafsson, K., & Dallner, M. (2002). Work environment and health in different types of temporary

jobs. European Journal of Work and Organizational Psychology, 11 , 151�/175.

Ashford, S. J., Lee, C., & Bobko, P. (1989). Content, causes and consequences of job insecurity: A theory-based

measure and substantive test. Academy of Management Journal , 32 , 803�/829.

Beard, K. M., & Edwards, J. R. (1995). Employees at risk: Contingent work and the psychological experience of

contingent workers. In C. L. Cooper, & D. M. Rousseau (Eds), Trends in organizational behaviour (pp. 109�/

126). Chichester, UK: John Wiley.

Beehr, T. A., Walsh, J. T., & Taber, T. D. (1976). Relationship of stress to individually and organizationally valued

states: Higher order needs as a moderator. Journal of Applied Psychology, 61 , 41�/47.

Benach, J., Gimeno, D., & Benavides, F. G. (2002). Types of employment and health in the European Union . Dublin:

European Foundation for the Improvement of Living and Working Conditions.

Benavides, F. G., Benach, J., Diez-Roux, A. V., & Roman, C. (2000). How do types of employment relate to health

indicators? Findings from the second European Survey on Working Conditions. Journal of Epidemiology and

Community Health , 54 , 494�/501.

Bernhard, C., & Sverke, M. (2003). Work attitudes, role stress and health among different types of contingent

workers in the Swedish health care sector. Research and Practice in Human Resource Management , 11 , 1�/16.

Blau, P. M. (1964). Exchange and power in social life . New York: Wiley.

Claes, R., De Witte, H., Schalk, R., Guest, D., Isaksson, K., Krausz, M., et al. (2002). Het psychologisch contract

van vaste en tijdelijke werknemers (The psychological contract of permanent and temporary workers). Gedrag en

Organisatie , 15 , 436�/455.

Cohen, S., Haberfeld, L., & Ferber, G. (1993). Temporary help service workers: Employment characteristics and

age determination. Industrial Relations , 32 , 272�/287.

Conway, N., & Briner, R. B. (2002). Full-time versus part-time employees: Understanding the links between work

status, the psychological contract, and attitudes. Journal of Vocational Behavior , 61 , 279�/301.

Corral, A., & Isusi, I. (2004). Part-time work in Europe . Dublin: European Foundation for the Improvement of

Living and Working Conditions.

Davis-Blake, A., & Uzzi, B. (1993). Determinants of employment externalisation: A study of temporary workers

and independent contractors. Administrative Science Quarterly , 38 , 195�/223.

De Cuyper, N., Isaksson, K., & De Witte, H. (Eds). (2005). Employment contracts and well-being among European

workers. Review of the literature . Aldershot: Ashgate.

de Jonge, J., Dorman, C., Janssen, P. P. M., Dollard, M. F., Landeweerd, J. A., & Nijhuis, F. J. N. (2001). Testing

reciprocal relationships between job characteristics and psychological well-being: A cross lagged structural

equation model. Journal of Occupational and Organizational Psychology, 74 , 29�/46.

de Lange, A. H., Taris, T. W., Kompier, M. A. J., Houtman, I., & Bongers, P. M. (2003). ‘The very best of the

Millenium’: Longitudinal research and the demand-control-(support) model. Journal of Occupational Health

Psychology, 8 , 282�/305.

De Witte, H., & Naswall, K. (2003). Objective versus subjective job insecurity: Consequences of temporary work

for job satisfaction and organizational commitment in four European countries. Economic and Industrial

Democracy, 24 , 149�/188.

Goldberg, D. (1979). Manual of the General Health Questionnaire . Windsor: National Foundation for Educational

Research.

Goudswaard, A., & Andries, F. (2002). Employment status and working conditions . Luxembourg: Office for Official

Publications of the European Communities.

316 C. Bernhard-Oettel et al.

Goudswaard, A., & de Nanteuil, M. (2000). Flexibility and working conditions: A European bibliographical review.

Dublin: European Foundation for the Improvement of Living and Working Conditions.

Gustafsson, S., Kenjoh, E., & Wetzels, C. (2001). Employment choices and pay differences between non-standard and

standard work in Britain, Germany, Netherlands and Sweden . Amsterdam: Tinbergen Institute.

Hackman, J. R., & Oldham, G. R. (1975). Development of a job diagnostic survey. Journal of Applied Psychology,

60 , 159�/170.

Hartley, J., Jacobson, D., Klandermans, B., & Van Vuuren, T. (1991). Job insecurity. Coping with jobs at risk .

Thousand Oaks, CA: Sage.

Hellgren, J. (2003). The times they are a-changin’ . Stockholm: Stockholm University.

House, R. J., & Rizzo, J. R. (1972). Role conflict and ambiguity as critical variables in a model of organizational

behavior. Organizational Behavior and Human Performance , 7 , 467�/505.

Isaksson, K. S., & Bellaagh, K. (2002). Health problems and quitting among female ‘temps’. European Journal of

Work and Organizational Psychology, 11 , 27�/45.

Isaksson, K., Bernhard, C., & Gustafsson, K. (2003). Betydelsen av anstallningsform och tillit till ledningen for

personalens halsa inom varden [The importance of employment contract and trust in management for health

care workers’ well-being]. Arbetsmarknad & Arbetsliv, 9 , 95�/106.

Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job redesign.

Administrative Science Quarterly, 24 , 285�/308.

Kinnunen, U., & Natti, J. (1994). Job insecurity in Finland: Antecedents and consequences. European Journal of

Work and Organizational Psychology, 4 , 297�/321.

Klein Hesselink, D. J., & Van Vuuren, T. (1999). Job flexibility and job insecurity: The Dutch case. European

Journal of Work and Organisational Psychology, 8 , 273�/293.

Krausz, M., Brandwein, T., & Fox, S. (1995). Work attitudes and emotional responses of permanent, involuntary,

and voluntary temporary help employees: An exploratory study. Applied Psychology: An International Review, 44 ,

217�/232.

Krausz, M., Sagie, A., & Bidermann, Y. (2000). Actual and preferred work schedules and scheduling control as

determinants of job-related attitudes. Journal of Vocational Behavior , 56 , 1�/11.

Letourneux, V. (1998). Precarious employment and working conditions in the European Union . Dublin: European

Foundation for the Improvement of Living and Working Conditions.

Martens, M. F. J., Nijhuis, F. J. N., van Boxtel, M. P. J., & Knottnerus, J. A. (1999). Flexible work schedules and

mental physical health. A study of a working population with non-traditional working hours. Journal of

Organizational Behavior , 20 , 35�/46.

McLean Parks, J., Kidder, D. L., & Gallagher, D. G. (1998). Fitting square pegs into round holes: Mapping the

domain of contingent work arrangements onto the psychological contract. Journal of Organizational Behavior ,

19 , 697�/730.

Morrison, D., Payne, R. L., & Wall, T. D. (2003). Is a job a viable unit of analysis? A multilevel analysis of demand-

control-support models. Journal of Occupational Health Psychology, 8 , 209�/219.

Organisation for Economic Co-operation and Development [OECD] (2002). OECD employment outlook . Paris:

OECD.

Paoli, P. (1996). Working conditions in the European Union (Summary) . Dublin: European Foundation for the

Improvement of Living and Working Conditions.

Paoli, P., & Merllie, D. (2001). Third European Survey on Working Conditions 2000 . Dublin: European Foundation

for the Improvement of Living and Working Conditions.

Reilly, P. A. (1998). Balancing flexibility �/ meeting the interests of employer and employee. European Journal of

Work and Organizational Psychology, 7 , 7�/22.

Richter, P., Hemman, E., Merboth, H., Fritz, S., Hansgen, C., & Rudolf, M. (2000). Das Erleben von

Arbeitsintensitat und Tatigkeitsspielraum �/ Entwicklung und Validierung eines Fragebogens zur orientierenden

Analyse (FIT) [Perception of work intensity and control*/development and validation of a questionaire].

Zeitschrift fuer Arbeits- und Organisationspychologie , 44 , 129�/139.

Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health

Psychology, 1 , 27�/41.

Sparrow, P., & Cooper, C. L. (2003). The employment relationship: Key challenges for HR . Oxford: Butterworth-

Heinemann.

Strazdins, L., D’Souza, R. M., Lim, L. L.-Y., Broom, D. H., & Rodgers, B. (2004). Job strain, job insecurity, and

health: Rethinking the relationship. Journal of Occupational Health Psychology, 9 , 296�/305.

Sverke, M., Gallagher, D. G., & Hellgren, J. (2000). Alternative work arrangements: Job stress, well-being, and

work attitudes among employees with different employment contracts. In K. Isaksson, L. Hogstedt, C. Eriksson,

& T. Theorell (Eds), Health effects of the new labour market (pp. 145�/167). New York: Plenum.

Employment contracts and health complaints 317

Sverke, M., Hellgren, J., & Naswall, K. (2002). No security: A meta-analysis and review of job insecurity and its

consequences. Journal of Occupational Health Psychology, 7 , 242�/264.

Sverke, M., & Sjoberg, A. (1994). Dual commitment to company and union in Sweden: An examination of

predictors and taxonomic split methods. Economic and Industrial Democracy , 15 , 531�/564.

Virtanen, M. (2003). Temporary employment and health . Helsinki: Finnish Institute of Occupational Health.

Virtanen, M., Kivimaki, M., Elovainio, M., Vahtera, J., & Cooper, C. L. (2001). Contingent employment, health

and sickness absence. Scandinavian Journal of Work. Environment & Health , 27 , 365�/372.

Virtanen, P., Vahtera, M., Kivimaki, J., & Pentii, J. F. (2002). Employment security and health. Journal of

Epidemiological Community Health , 56 , 569�/574.

Walsh, J., Taber, T. D., & Beehr, T. A. (1980). An integrated model of perceived job characteristics. Organizational

Behavior and Human Performance , 25 , 252�/267.

Wikman, A. (2002). Temporara kontrakt och inlasningseffekter (Temporary contracts and being locked-in effects).

Arbetsliv i Omvandling , 4.

Wikman, A., Andersson, A., & Bastin, M. (1998). Nya relationer i arbetslivet (New relations in working life) .

Stockholm: National Institute for Working Life.

318 C. Bernhard-Oettel et al.